EMOTION REGULATION IN EVERYDAY LIFE 1
Emotion Regulation in Everyday Life:
Mapping Global Self-Reports to Daily Processes
Peter Koval1,2, Elise Kalokerinos1, Katharine Greenaway1, Hayley Medland1, Peter Kuppens2,
John B. Nezlek3,4, Jordan D. X. Hinton1,5, & James J. Gross6
1 University of Melbourne, Australia
2 KU Leuven, Belgium
3 SWPS University of Social Sciences and Humanities at Poznań, Poland
4 College of William and Mary, USA
5 Australian Catholic University, Australia
6 Stanford University, USA
Author Note
This research was supported by a Discovery Project grant (DP160102252), Discovery Early
Career Researcher Awards (DE190100203; DE180100352) and a Future Fellowship
(FT190100300) from the Australian Research Council, a Horizon 2020 Marie Skłodowska-
Curie fellowship (704298) and by a KU Leuven Research Council grant (C14/19/054).
Correspondence concerning this article should be addressed to Peter Koval, Melbourne
School of Psychological Sciences, The University of Melbourne, Parkville VIC 3010,
Australia; email: [email protected] EMOTION REGULATION IN EVERYDAY LIFE 2
Abstract
Recent theory conceptualizes emotion regulation as a dynamic process occurring across three stages: (i) identifying the need to regulate, (ii) selecting a strategy, and (iii) implementing that strategy to change an emotional state. Despite its dynamic nature, emotion regulation is typically assessed using static global self-report questionnaires that ask people to reflect on their general use of certain strategies. It is unclear how well these global self-reports capture the identification, selection, and implementation stages outlined in the theory. To address this issue, we examined how global self-report measures correspond with the three stages of emotion regulation, modelled using daily life data. We analyzed data from nine daily diary and experience sampling studies (total N=1,097), in which participants provided daily and global self-reports of cognitive reappraisal, expressive suppression, and rumination. Results across studies revealed weak to moderate correlations between global self-reports and individual differences in strategy selection in daily life. We also found evidence that suppression and rumination global self-reports capture identification processes, and rumination global self-reports capture implementation processes. Our findings suggest that global self-reports do not only assess trait strategy selection but may also reflect individual differences in identification and implementation of emotion-regulation strategies in daily life. EMOTION REGULATION IN EVERYDAY LIFE 3
A rapidly growing literature on emotion regulation has documented how people seek to influence their emotions. Early work focused on developing taxonomies of emotion- regulation strategies (e.g., Gross, 1998; Koole, 2009; Parkinson & Totterdell, 1999) and understanding each strategy’s distinct affective consequences (e.g., Aldao et al., 2010; Webb et al., 2012a). In contrast, recent theory has conceptualized emotion regulation as a dynamic, multi-stage process (e.g., Gross, 2015; Webb et al., 2012b), highlighting the importance of understanding both the antecedents and consequences of regulation strategies (e.g., Sheppes et al., 2014). Yet, our measurement of emotion regulation has not entirely kept pace with theoretical advances. Global self-report questionnaires continue to be very common, if not the most widely used, measures of emotion regulation in daily life (e.g., John & Eng, 2014).1
Although global self-reports are routinely assumed to assess people’s typical use of emotion- regulation strategies, it remains unclear how well they do so, and whether they also index more dynamic aspects of the daily regulation process.
Research using daily diary (DD; e.g., Nezlek & Kuppens, 2008) and experience sampling methods (ESM; e.g., Brans et al., 2013) provides a means to capture the dynamic, multi-stage processes theorized to underpin everyday emotion regulation (Aldao et al., 2015;
Colombo et al., 2020). Thus, studies combining such daily life methods with global questionnaires are needed to determine how static global self-reports align with the various aspects of emotion regulation dynamics outlined in recent theoretical accounts. To this end, we investigate how global self-reports of three widely-studied emotion-regulation strategies correspond with the dynamics of these strategies in daily life, using data from nine DD and
ESM studies.
1 We recognize that a substantial body of research has adopted a different approach to studying emotion regulation, in which participants are experimentally instructed to use particular regulation strategies, mostly in the lab (e.g., Morawetz et al., 2017; Webb et al., 2012a). However, our focus in the current paper is on understanding the processes involved in spontaneous (i.e., uninstructed) emotion regulation in daily life. EMOTION REGULATION IN EVERYDAY LIFE 4
Emotion Regulation as a Multi-Stage Dynamic Process
Recent theoretical accounts of emotion regulation emphasize its dynamic, iterative, and cyclical nature, including bidirectional links with emotions (e.g., Gross, 2015; Tamir,
2021; Webb et al., 2012b). In particular, Gross’s (2015) extended process model proposes three sequential stages: (i) an identification stage, during which people appraise their current emotions as “good” or “bad” for them and activate a goal to regulate; (ii) a selection stage, during which people choose which regulation strategies to use; and (iii) an implementation stage, involving the deployment of selected strategies to influence emotions.
A major implication of these accounts (e.g., Gross, 2015) is that a full understanding of emotion regulation requires us to move beyond the singular focus on which strategies people select. Researchers must also consider when (e.g., Haines et al., 2016) and how effectively (e.g., Ford et al., 2017) people deploy their chosen emotion-regulation strategies
(see also Sheppes, 2020). However, only a handful of empirical studies have attempted to operationalize emotion regulation as a dynamic multi-stage process, incorporating both the antecedents and consequences of regulation (e.g., Lennarz et al., 2019; Livingstone &
Isaacowitz, 2019; Pavani et al., 2017).
Global Self-Reports of Emotion Regulation
The dominant approach to studying emotion regulation in daily life relies on global
(aka. “trait”) questionnaires (for reviews, see John & Eng, 2014; Naragon-Gainey et al.,
2017). These measures ask respondents to report the extent to which they habitually or
“typically” use a range of regulation strategies in their daily lives (e.g., De France &
Hollenstein, 2017; Garnefski et al., 2001; Gratz & Roemer, 2004; Gross & John, 2003;
Nolen-Hoeksema & Morrow, 1991). Such global self-reports are commonly assumed to measure individual differences in people’s habitual selection of emotion-regulation strategies
(John & Eng, 2014; McMahon & Naragon-Gainey, 2020). For example, in their meta- EMOTION REGULATION IN EVERYDAY LIFE 5 analysis of associations between global self-reported emotion regulation and psychopathology, Aldao et al. (2010) argue that “self-report scales typically measure dispositional tendencies toward certain emotion-regulation strategies, and thus supposedly assess what participants do across time and different contexts” (p. 219). Yet, as Naragon-
Gainey et al. (2017) point out, it remains unclear how self-reported habitual strategy use maps onto how people identify, select, and implement regulation strategies in daily life.
We argue that although global self-report measures are inherently static, they may nevertheless index some dynamic aspects of everyday emotion regulation, such as when and how effectively people tend to use specific emotion-regulation strategies. To test such a claim requires methods capable of capturing these dynamic aspects of emotion regulation. As we outline below, daily life methods offer an ideal candidate.
Daily Self-Reports of Emotion Regulation
Daily life methods provide a window into emotion regulation as a dynamic process by obtaining intensive momentary self-reports of emotion and its regulation in everyday life
(e.g., Colombo et al., 2020). Although nascent, research using daily life methods to study emotion regulation is growing (e.g., Blanke et al., 2019; Brans et al., 2013; Heiy & Cheavens,
2014; Medland et al., 2020; Nezlek & Kuppens, 2008; Troy et al., 2019). Measures in this literature tend to index emotion-regulation strategy selection, in that they document the which strategies people use in everyday life. Findings to date reveal that strategy use in daily life varies considerably both within and between individuals. Whereas the within-person variance in daily self-reports reflects state selection of emotion-regulation strategies, the between- person variance in daily self-reports reflects trait strategy selection (assuming daily life contexts are representatively sampled; Hamaker et al., 2007; Hamaker et al., 2017).
Yet, the capacity for daily life methods to assess emotion regulation dynamics goes beyond the strategy selection stage. Specifically, daily self-reports of emotion regulation can EMOTION REGULATION IN EVERYDAY LIFE 6 be used to model individual differences in within-person contingencies between state affect and regulation strategy use, which are (at least partly) driven by processes occurring in the identification and implementation stages of regulation (Gross, 2015). For the identification stage, we draw on the fact that identifying a need to regulate relies on a representation of the current affective state (Gross et al., 2019). Thus, we propose that one way to operationalize identification is using the degree to which state strategy selection is predicted by recent affective experience. In contrast, the implementation stage is where regulation processes reach their target, as strategies influence how emotions unfold over time (Gross et al., 2019).
Thus, the extent to which state strategy selection predicts change in subsequent affective experience is one way to operationalize the implementation stage.
These preliminary operationalizations of the identification and implementation stages are undoubtedly imperfect given the highly complex and iterative nature of emotion regulation (Gross, 2015; Gross et al., 2019). Yet, we propose that these operationalizations provide a way to increase theoretical insights obtained from daily self-reports of emotion regulation-strategy use. In addition to providing a more nuanced test of theory, these operationalizations of daily emotion regulation stages can be used to inform understanding of existing global measures of emotion regulation. Specifically, in the present research we investigated how global self-reports correlate with individual differences in the identification, selection, and implementation stages of emotion regulation, operationalized using daily self- reports of emotion regulation and affect.
Correspondence Between Global and Daily Self-Reports of Emotion Regulation
In the following paragraphs, we outline research and theory relevant to the correspondence between global-self reports and each of the three stages of emotion regulation in daily life. We begin by discussing how global self-reports correspond with individual differences in daily strategy selection, because we argue that researchers have EMOTION REGULATION IN EVERYDAY LIFE 7 commonly (albeit implicitly) assumed that global self-report scales assess trait strategy selection. That is, global self-reports are generally assumed to give information about which strategies people habitually select to regulate their emotions. Next, we turn our attention to correspondence between global self-reports and individual differences in identification and implementation processes in daily life. That is, we explore the possibility that, in addition to targeting strategy selection, global self-reports also, in part, index when people use certain strategies, and how using those strategies influences subsequent affect. Together, these tests of global-daily self-report correspondence aim to yield greater insight into common measures of emotion regulation, and which stage(s) of emotion regulation they assess.
Selection Correspondence
Previous research on the convergence between global and daily self-reports of emotion regulation has been limited to examining how global self-report scales correlate with between-person variance in daily self-reports of emotion-regulation strategy selection (e.g.,
Brockman et al., 2017; McMahon & Naragon-Gainey, 2020; Schwartz et al., 1999). We refer to this correlation between global and daily self-reports of strategy selection as selection correspondence. Two existing theoretical frameworks support distinct predictions about the likely strength of selection correspondence.
According to whole trait theory (Fleeson & Jayawickreme, 2015), traits can be defined as density distributions of states, such that the mean of repeatedly sampled states represents a person’s typical standing along a trait dimension. Since global self-reports also capture people’s (perceived) typical trait level, they should correlate relatively strongly with mean daily self-reports. Supporting this prediction, Fleeson and Gallagher (2009) found that mean state self-reports of the Big Five personality dimensions correlate at r = .40 to .60 with global self-reports of the same dimensions (see also Augustine & Larsen, 2012; Finnigan &
Vazire, 2018). If emotion regulation can be subsumed under whole trait theory, we would EMOTION REGULATION IN EVERYDAY LIFE 8 expect to find relatively strong selection correspondence, in line with how global self-reports of emotion-regulation strategies have been interpreted in the literature to date. However, as whole trait theory was developed to explain the structure of broad personality dimensions it may be less applicable to narrower constructs, such as emotion-regulation strategies.
In contrast, Robinson and Clore’s (2002) accessibility model of emotional self- report proposes that people draw on different knowledge when reporting global versus momentary feelings. Momentary reports are a direct read-out of experiential knowledge, which is highly contextualized, embodied, and decays rapidly in episodic memory. In contrast, because feelings are context-dependent and fleeting, people must rely on abstract semantic knowledge about the self when providing global self-reports of how they typically feel (Robinson & Clore, 2002; see also Conner & Barrett, 2012). Thus, the accessibility model of emotional self-report predicts weak or no correspondence between global and momentary measures of feelings. Given that daily emotion regulation may be equally dynamic, fluctuating in response to affect (e.g., Feldman & Freitas, 2021; Pavani et al.,
2017), goals (e.g., English et al., 2017; Kalokerinos et al., 2017) and other situational factors
(e.g., Sheppes et al., 2014; Young & Suri, 2020), global and momentary reports of emotion regulation-strategies may similarly derive from different sources of information. From this perspective, we would predict weak selection correspondence. Yet, unlike feelings, emotion regulation-strategies involve deliberate behavioral or cognitive efforts and may thus be easier to encode as discrete “events” for later retrieval. If this were the case, Robinson and Clore’s
(2002) model may be less applicable to emotion-regulation strategies.
Previous studies provide mixed evidence regarding the strength of selection correspondence, with some studies finding no reliable convergence between daily and global ratings of emotion-regulation strategy selection (e.g., Brockman et al., 2017; Schwartz et al.,
1999), and others reporting reliable, yet relatively weak, correspondence (rs ≤ .35; e.g., Ford EMOTION REGULATION IN EVERYDAY LIFE 9 et al., 2017; Kircanski et al., 2015; Pasyugina et al., 2015; Peters et al., 2020). A recent study by McMahon and Naragon-Gainey (2020), which examined correlations among global and daily self-reports of 10 emotion-regulation strategies in two daily diary studies, represents the most comprehensive test of selection correspondence to date. Despite assessing daily regulation strategy selection in a highly relevant context (i.e., intense emotional episodes),
McMahon and Naragon-Gainey found mixed support for strategy-specific selection correspondence (rs = -.03 to .64). Further, global self-reports of each emotion-regulation strategy tended to correlate as strongly with daily reports of other strategies as with daily reports of the same strategy, suggesting that global measures show low discriminant validity.
Taken together, previous research provides mixed evidence regarding how well global self-report measures capture people’s typical selection of emotion-regulation strategies in daily life. In the present research, we provide a robust test of selection correspondence for three widely-studied regulation strategies across nine diverse diary and ESM studies, more than quadrupling the sample size of previous investigations of selection correspondence
(McMahon & Naragon-Gainey, 2020).
Identification Correspondence
The current study goes beyond previous tests of the ecological validity of global self- report measures of emotion regulation by examining how global self-reports correlate with two dynamic aspects of daily emotion regulation. First, we examine how global self-reports correlate with the within-person contingency between daily strategy use and preceding affective experience, which we propose as a measure of identification correspondence.
Emotion regulation occurs when people evaluate their current emotion as requiring adjustment and thus should vary as a function of affective experience.
Previous studies have demonstrated that daily use of emotion-regulation strategies is contingent on recent affective experience (Brans et al., 2013) and that such dynamics vary as EMOTION REGULATION IN EVERYDAY LIFE 10 a function of personality (Pavani et al., 2017). However, to our knowledge, the current research is the first to investigate how individual differences in identification (operationalized as the contingency of state emotion regulation on preceding affect) correlate with global self- reports.
We reasoned that identification correspondence may emerge because many global self-report emotion regulation questionnaires bake affective input into their items or scale instructions. For instance, the Ruminative Responses Scale (RRS; Nolen-Hoeksema &
Morrow, 1991) instructs participants to indicate how much they engage in rumination when feeling “down, sad, or depressed”. Similarly, several items in the Emotion Regulation
Questionnaire (ERQ; Gross & John, 2003) explicitly invoke emotional antecedents (e.g.,
“when I’m feeling negative emotions…”) of strategy selection.
Two recent studies are consistent with this claim that global emotion regulation scales may index the identification of a need to use regulation strategies. Tamir et al. (2019) recently demonstrated that the ERQ reappraisal subscale partly measures pro-hedonic emotion goals (i.e., wanting to decrease negative affect or increase positive affect), which become activated during the identification stage (Gross, 2015). Similarly, self-reported habitual expressive suppression (assessed with the ERQ) has been found to correlate more strongly with suppression in daily life in the context of more (vs. less) intense negative emotions (Peters et al., 2020). This suggests that global self-report measures, such as the
ERQ, may capture the tendency to use regulation strategies specifically when there is a greater need to regulate, rather than across all contexts (Peters et al., 2020).
Implementation Correspondence
Second, we examine how global self-reports relate to the within-person contingency between daily strategy use and subsequent affective experience, which we label implementation correspondence. Implementation is the final stage of emotion regulation, in EMOTION REGULATION IN EVERYDAY LIFE 11 which a person applies selected emotion-regulation strategies with the aim of influencing their subsequent emotion (Gross, 2015). Thus, although the affective consequences of emotion regulation cannot be completely divorced from strategy selection (i.e., regulation strategies differ in their affective consequences; Webb et al., 2012a), we consider the impact of using a particular regulation strategy on affective experience to be driven primarily by how it is implemented.
Previous research has documented that daily use of emotion-regulation strategies reliably impacts subsequent affective experience (e.g., Brans et al., 2013) and that these implementation effects differ substantially between individuals (e.g., Pavani et al., 2017; Pe et al., 2013; Troy et al., 2019). Although we are not aware of any studies that directly investigate whether individual differences in strategy implementation correlate with global self-reports of emotion-regulation strategies, there are theoretical reasons to suppose such implementation correspondence may occur. Specifically, theories of personality dynamics and development suggest that associative learning processes may increase people’s tendency to engage in behaviours that are rewarding when implemented (Wrzus & Roberts, 2017).
That is, people may be more likely to use a regulation strategy if implementing that strategy results in greater desired emotional change. Alternatively, people may simply be more likely to remember using a regulation strategy to the extent that it has any (desired or undesired) affective consequences (Kensinger, 2009).
In line with this reasoning, Ford et al. (2017) found that global self-reports of cognitive reappraisal correlated equally strongly with a daily measure of perceived reappraisal success (r = .23) as with mean daily use of reappraisal (r = .18), suggesting that global self-report measures of reappraisal may not only index habitual selection of this strategy, but also how effectively it is implemented. Other studies have similarly found that global self-reports of reappraisal are associated with reappraisal effectiveness (e.g., Kanske et EMOTION REGULATION IN EVERYDAY LIFE 12 al., 2015; McRae et al., 2012), suggesting that global self-reports of emotion regulation may index individual differences in strategy implementation, as well as selection.
The Present Research
Our aim was to investigate how global self-reports of emotion regulation map onto the identification, selection, and implementation of regulation strategies in daily life (Gross,
2015). Given the prevalence of global self-report scales, this mapping is critical to theoretically situating and interpreting a large portion of research on emotion regulation. To test for selection correspondence, we estimated correlations between global self-reports and individual differences in the between-person (mean) component of daily self-reports of emotion regulation. We tested for identification correspondence by estimating how global self-reports correlate with the degree to which daily emotion regulation is contingent on preceding positive and negative affective experience. Finally, to investigate implementation correspondence, we modelled how global self-reports of emotion regulation correspond with the impact of daily emotion regulation on subsequent positive and negative affective experience. In each case, we examined correspondence both within and across regulation strategies to provide evidence of convergent and discriminant associations, respectively, among daily and global measures.
We addressed these questions using data from nine studies in which participants completed global and daily self-reports of three widely-studied and distinct emotion- regulation strategies: (i) cognitive reappraisal, a cognitive strategy involving the reframing of emotional situations to alter their emotional impact; (ii) expressive suppression, a behavioral strategy involving the inhibition of emotionally expressive reactions; and (iii) rumination, a cognitive strategy involving repetitive and passive engagement with one’s emotions.2
2 Consistent with most research on rumination, we focused on the maladaptive brooding component of rumination (Treynor et al., 2003; see also Trapnell & Campbell, 1999). EMOTION REGULATION IN EVERYDAY LIFE 13
Participants reported their habitual use of these regulation strategies at baseline using global questionnaires, and subsequently completed DD or ESM reports of their state use of each strategy 1-10 times per day for 7-21 consecutive days. We report separate analyses per study as well as meta-analyses to summarize evidence across all nine studies.
Method
We analyzed data from nine daily life studies (Total N=1,097). Six of these studies
(two DD and four ESM studies) asked participants to report on their use of emotion- regulation strategies across everyday contexts. The other three ESM studies assessed emotion-regulation strategies in response to emotional or stressful events—we label these
“ESM-Event” studies. Combining data from these nine studies with diverse designs and samples served to maximize statistical power and increase the generalizability of our findings. Below, we report analyses for each DD, ESM, and ESM-Event study followed by meta-analyses summarizing results across all relevant studies.
All studies received appropriate ethical clearance (or were exempt from ethics review) and have been previously analyzed to answer research questions distinct from those addressed in this paper (see Table S1 in the supplemental materials for more information about each study). Below, we report only the measures relevant to the current research.
Participants and Procedure
Participants were undergraduates (6 studies), general community members (2 studies), or Amazon MTurk workers (1 study) recruited in Australia, Belgium or the US. In all studies, participants completed both global and daily self-report measures of cognitive reappraisal, expressive suppression, and rumination. We followed exclusion criteria applied in previous analyses of each dataset. Further details of participants and procedure for each study are summarized in Table S1 in the supplemental materials. EMOTION REGULATION IN EVERYDAY LIFE 14
Materials and Measures
Global Self-Reports of Emotion-Regulation strategies
Global reappraisal and suppression use were assessed with Gross and John’s (2003)
ERQ, comprising six reappraisal items (e.g., “When I want to feel less negative emotion
(such as sadness or anger), I change what I’m thinking about”), and four suppression items
(e.g., “When I am feeling negative emotions, I make sure not to express them”) rated from 1
(strongly disagree) to 7 (strongly agree).
In all but two studies, global rumination was measured with the brooding subscale of the RRS (Treynor et al., 2003), comprising five items (e.g., “Think ‘Why do I always react this way?’”) rated on a scale from 1 (almost never) to 4 (almost always). Two studies (DD
Study 2, ESM Study 3) assessed global rumination using the rumination subscale of Trapnell and Campbell’s (1999) Reflection-Rumination Questionnaire, comprising 12 statements (e.g.,
“I never ruminate or dwell on myself for very long”; reverse-scored) rated from 1 (strongly disagree) to 5 (strongly agree).
Descriptive statistics and reliability estimates for the global emotion regulation scales are reported in Table S4 in the supplemental materials.
Daily Self-Reports of Emotion-Regulation strategies
As is typical for DD and ESM studies, emotion-regulation strategies were measured with single items in most studies. Specifically, single-item measures were used in six studies to measure reappraisal (e.g., “Have you looked at the cause of your feelings from another perspective?”), in seven studies to assess suppression (e.g., “I was careful not to express my emotions to others”), and in five studies to measure rumination (e.g., “I ruminated or dwelled on the event or my emotions”). The remaining studies included 2- or 3-item measures of daily emotion-regulation strategies (see Table S2 in the supplemental materials). Descriptive EMOTION REGULATION IN EVERYDAY LIFE 15 statistics (and reliability estimates for multi-item measures) for daily self-report measures of emotion-regulation strategies are reported in Table S5 in the supplemental materials.
Daily Self-Reports of Affect
In all but one study (ESM-Event Study 3), momentary negative affect (NA) and positive affect (PA) were assessed with two or more items including at least one low-arousal
(e.g., sad, relaxed) and one high-arousal (e.g., angry, happy) adjective. In ESM-Event Study
3, NA and PA were assessed using single items (see Table S3 in the supplemental materials for complete NA and PA items in all ESM and ESM-Event studies). Multilevel reliability
(Cronbach’s alpha) for daily NA and PA scales are reported in Table S6 in the supplemental materials. We do not provide details of affect measures in the two DD studies as we only conducted analyses including affect for ESM and ESM-Event studies (see below).
Data Analyses
Power and Sample Size
The current project involved secondary data analysis of existing DD and ESM studies. Thus, sample sizes for each study were determined according to various theoretical and pragmatic criteria. As such, some individual studies (e.g., ESM Study 3, N = 46) are likely underpowered to test our research questions. We therefore conducted meta-analyses
(see below) to maximize statistical power and synthesize findings across studies. To ensure that our meta-analyses were sufficiently powered, we ran post-hoc power analyses (see Table
S14 in the supplemental materials).
Multilevel Structural Equation Models (MSEM)
To account for the nested data structure (DD/ESM surveys nested within participants), we analyzed data using multilevel structural equation modeling (MSEM) in
Mplus version 8.5 (L.K. Muthén & Muthén, 1998-2017). MSEM uses latent centering to decompose observed multilevel outcomes into latent within-person (state) and between- EMOTION REGULATION IN EVERYDAY LIFE 16 person (trait) components and allowed us to specify models with both regressions and covariances among multiple outcomes. Due to the inclusion of random slopes for predictors with missing data, using maximum likelihood estimation for our analyses would have required high-dimensional numerical integration, which is computationally intensive and leads to poor model convergence. Thus, we used Bayesian estimation3 as recommended by
Asparouhov and Muthen (2019). Bayesian estimation also allowed us to obtain standardized parameter estimates for subsequent meta-analysis. We report standardized estimates and 95%
Bayesian credible intervals (CIs) for all parameters, which we considered as “statistically significant” when their 95% CIs did not include zero.
Figure 1 illustrates the MSEM fitted to each dataset. The left panel shows the decomposition of observed DD/ESM variables into latent within-person and between-person components. Across all nine studies, we tested for selection correspondence by correlating global self-reports with the between-person component of daily self-reports for each emotion- regulation strategy (see Figure 1, top-right panel).
We modelled identification and implementation slopes using the within-person state components of DD/ESM measures (see Figure 1, bottom-right panel). We only fitted this within-person model to the seven ESM/ESM-Event studies because we reasoned that identification and implementation processes unfold within, but not across, days. This aligns with the finding that most daily emotional episodes have a duration of 1-2 hours (Verduyn et al., 2009). In the two DD studies it would have only been possible to model how emotion- regulation strategy use on a given day predicts affect the next day (or vice versa), which we did not consider to appropriately reflect the timescale of identification and implementation dynamics.
3 We used Mplus’s default non-informative Bayesian priors ensuring that our results are asymptotically equivalent to those obtained under maximum likelihood (Zyphur & Oswald, 2015). To obtain stable parameter estimates, we specified a minimum of 20,000 Bayesian iterations before checking for model convergence, which was defined as a posterior scale reduction value below 1.05. EMOTION REGULATION IN EVERYDAY LIFE 17
Within-person identification slopes were estimated by regressing each emotion- regulation strategy at time t onto affect at t –1, while controlling for regulation strategies at t
–1. Thus, within-person identification slopes represent the extent to which state emotion regulation is contingent upon preceding affect. We tested for identification correspondence by correlating global self-reports of emotion regulation with (random) identification slopes at the between-person level. Within-person implementation slopes were estimated by regressing affect at time t onto emotion-regulation strategy use “since the last survey” reported at time t, controlling for affect at t –1. Within-person implementation slopes therefore represent how the state use of each emotion-regulation strategy is associated with change in affect over time. We tested for implementation correspondence by correlating global emotion regulation scores with (random) implementation slopes at the between-person level.
All within-person regression paths (including autoregressive effects) were estimated as random slopes and allowed to freely covary with all other parameters at the between- person level (see top-right panel in Figure 1).
Meta-Analyses
We extracted standardized covariances (i.e., correlations) and regression coefficients from each study (Peterson & Brown, 2005) and synthesized these using random-effects meta- analyses using the metafor R-package (Viechtbauer, 2010). Meta-analyses for selection correspondence included estimates from all nine studies, whereas those for identification and implementation correspondence used estimates from the seven ESM/ESM-Event studies. We also report meta-analytic estimates for average identification (Tables S8 and S9) and implementation (Tables S10 and S11) slopes in the supplemental materials.
Open Practices
All data and analysis scripts required to reproduce the analyses reported in this paper are available at https://osf.io/q5dz6/. EMOTION REGULATION IN EVERYDAY LIFE 18
Results
Selection Correspondence
Table 1 contains correlations among global self-reports and the between-person
(trait) component of daily self-reports, representing selection correspondence. Shaded cells contain within-strategy correlations among global and daily self-reports for each emotion- regulation strategy, and unshaded cells contain cross-strategy correlations (e.g., correlations between global reappraisal and daily rumination). We first present individual study results focusing on within-strategy correlations, before presenting meta-analytic findings across all studies for both within-strategy and cross-strategy correlations.
Individual Study Results
In the DD studies, within-strategy correlations representing selection correspondence for each strategy ranged from .24 to .56 and were all meaningfully non-zero
(i.e., 95% CIs did not include zero). For the ESM studies, results were more mixed: the within-strategy correlations representing selection correspondence ranged from .06 to .46, only some of which were reliably non-zero. Finally, for the ESM-Event studies, selection correspondence was weaker and less consistent, with within-strategy correlations ranging from –.22 to .34, although not all were reliably non-zero. In all three types of studies, rumination showed the strongest selection correspondence, followed by suppression and then reappraisal. EMOTION REGULATION IN EVERYDAY LIFE 19
Between
global global global
REAP SUPR RUM
"#$% "#$% $&& "#$% $&& "$&& ! #"#$% $"#$% !
"#$% '(%" "() NA→ Between ! ! ! $** )'%" $&& )'%" ! ! ! !! " )'%" # SUPR$ ! $&& $&& ! )'%% )'%" $$&&
"'( "'( $&& "'( "$&& ! #"'( $"'(
Observed Daily Self- REAP&! SUPR&! S1 RUM&! AFF&! $%&' Reports Within !$%&' !"#"$ !"#"$ REAP !"# REAP!
()'$ !"#"$ !"#"$ !"#"$ !"#"$ !"#"$ !()'$ !"#"$ Within !"" REAP! SUPR! RUM! AFF! SUPR!"# SUPR! &#$!% #$!% %!""
$)* (&%# !$)* %!"" !"" !"#"$ !"#"$ &(&%# RUM RUM !"# ! #&' !"" %!"" &#&' !&++ !"#"$ &++ !"#"$ AFF!"# AFF!
t –1 t
Figure 1. MSEM model diagram. Left panel shows decomposition of observed daily self-reports (grey rectangles) into latent within-person (white ellipses) and between-person (black #$!% ellipses) components. Bottom-right panel shows the within-person model: identification slopes (e.g., !!"" ) were estimated by regressing each regulation strategy at time t (reported as !"" ‘since the last survey’) onto state affect reported at time t – 1, controlling for state regulation strategy use at time t – 1; implementation slopes (e.g., "#$!%) were estimated by regressing #$!% state affect at time t onto state regulation strategies at time t, controlling for affect at time t – 1; autoregressive slopes (e.g., $#$!% ) were estimated using observed person-mean centered lagged variables. All within-person intercepts and slopes had random effects. Top-right panel shows the between-person model: all random effects were allowed to covary freely with each other and with observed global self-reports of each emotion-regulation strategy (e.g., %&'()&#$!%). EMOTION REGULATION IN EVERYDAY LIFE 20
Meta-Analyses
Results across all nine studies revealed that global self-reports showed reliable positive correlations with the between-person component of daily self-ratings for all three emotion-regulation strategies. Nevertheless, there were substantial differences among strategies, with the strongest selection correspondence for rumination (rmeta = .40, 95% CI
[.30 to .49]), followed by suppression (rmeta = .30, 95% CI [.19 to .42]), and the weakest selection correspondence for reappraisal (rmeta = .14, 95% CI [.01 to .28]).
We also found evidence of reliable cross-strategy correlations, in particular for global rumination. Specifically, global rumination correlated with the between-person components of daily reappraisal (rmeta = .13, 95% CI [.03 to .24]) and suppression (rmeta = .25,
95% CI [.15 to .35]) to a similar degree as global self-reports of reappraisal and suppression.
Additional Exploratory Analyses
We ran a series of exploratory analyses testing whether individuals with greater variability in daily self-reports of emotion regulation may be less accurate in their global self- reports. Meta-analyses of these models revealed that selection correspondence was not reliably moderated by within-person variability in daily self-reports for any of the regulation strategies (see Table S7 in the supplemental materials).
Identification Correspondence
Tables 2 and 3 contain within-strategy (shaded) and cross-strategy (unshaded) correlations between global self-reports and within-person NA and PA identification slopes, respectively. Positive correlations indicate that individuals with higher global self-reports tend to have more positive identification slopes, which reflect the degree to which increases in state regulation strategy use are contingent on preceding affect. We first briefly report average identification slopes for each strategy to aid in the interpretation of our key findings.
Then, first for NA and then for PA, we report the identification correspondence results for
EMOTION REGULATION IN EVERYDAY LIFE 21
each study, focusing on within-strategy correlations. Finally, we report meta-analytic findings
across all studies, including both within- and cross-strategy correlations.
Table 1 Correlations Among Global Self-Reports and the Trait-Component of Daily Self-Reports Representing Selection Correspondence Between-Person Component of Daily Self-Reports "#$% &'%" "'( Reappraisal (!! ) Suppression (!! ) Rumination (!! ) Global Self-Reports r 95% CI r 95% CI r 95% CI Reappraisal DD Study 1 (N = 114) .34 .15 to .55 .08 -.15 to .30 -.31 -.50 to -.10 DD Study 2 (N = 153) .38 .23 to .52 -.01 -.18 to .16 -.02 -.19 to .15 ESM Study 1 (N = 176) .20 .04 to .36 .06 -.11 to .22 .06 -.11 to .22 ESM Study 2 (N = 200) .06 -.09 to .21 -.02 -.17 to .13 .07 -.09 to .22 ESM Study 3 (N = 46) .10 -.46 to .64 -.01 -.57 to .53 .05 -.50 to .60 ESM Study 4 (N = 95) .08 -.17 to .34 -.13 -.37 to .12 -.07 -.32 to .19 ESM-Event Study 1 (N = 101) -.22 -.44 to .02 -.17 -.40 to .07 -.17 -.40 to .06 ESM-Event Study 2 (N = 100) -.05 -.29 to .20 .01 -.24 to .25 -.11 -.34 to .14 ESM-Event Study 3 (N = 112) .32 .11 to .51 .12 -.10 to .34 .19 -.03 to .40 Meta-Analysisa (N = 1097) .14 .01 to .28 -.01 -.06 to .06 -.03 -.13 to .07 Suppression DD Study 1 (N = 114) -.20 -.40 to .03 .24 .03 to .45 .06 -.15 to .28 DD Study 2 (N = 153) .03 -.14 to .19 .56 .45 to .68 .21 .04 to .36 ESM Study 1 (N = 176) .17 .01 to .33 .45 .31 to .57 .37 .22 to .51 ESM Study 2 (N = 200) .01 -.14 to .17 .18 .02 to .32 .07 -.09 to .22 ESM Study 3 (N = 46) .03 -.55 to .56 .21 -.37 to .70 .08 -.48 to .61 ESM Study 4 (N = 95) -.06 -.30 to .20 .26 .01 to .48 .01 -.23 to .27 ESM-Event Study 1 (N = 101) .00 -.23 to .25 .13 -.11 to .36 -.02 -.26 to .22 ESM-Event Study 2 (N = 100) .12 -.14 to .34 .28 .04 to .49 .23 -.01 to .45 ESM-Event Study 3 (N = 112) -.12 -.34 to .10 .22 .01 to .43 .03 -.19 to .25 Meta-Analysisa (N = 1097) .00 -.08 to .08 .30 .19 to .42 .13 .03 to .22 Rumination DD Study 1 (N = 114) -.13 -.34 to .09 -.05 -.26 to .18 .43 .24 to .60 DD Study 2 (N = 153) .14 -.03 to .30 .36 .20 to .50 .56 .44 to .67 ESM Study 1 (N = 176) .26 .11 to .41 .43 .30 to .57 .46 .32 to .58 ESM Study 2 (N = 200) .26 .12 to .40 .26 .12 to .40 .45 .32 to .57 ESM Study 3 (N = 46) -.21 -.69 to .38 .03 -.53 to .59 .11 -.46 to .63 ESM Study 4 (N = 95) .10 -.15 to .34 .26 .02 to .49 .31 .07 to .53 ESM-Event Study 1 (N = 101) .22 -.01 to .45 .29 .06 to .50 .26 .02 to .47 ESM-Event Study 2 (N = 100) .25 .01 to .46 .27 .04 to .50 .34 .11 to .54 ESM-Event Study 3 (N = 112) .13 -.09 to .34 .21 -.01 to .42 .24 .02 to .44 Meta-Analysisa (N = 1097) .13 .03 to .24 .25 .15 to .35 .40 .30 to .49 Note. Shaded cells contain within-strategy correlations representing selection correspondence for each emotion-regulation strategy; Estimates in bold have 95% CIs that do not include zero. a Meta-analyses based on Fisher z-transformed correlations with frequentist 95% CIs.
EMOTION REGULATION IN EVERYDAY LIFE 22
Average Identification Slopes
Meta-analyses of the average identification slopes revealed that NA predicted increases in state rumination whereas PA predicted decreases in state rumination. In contrast, neither state reappraisal nor suppression were consistently predicted by preceding NA or PA
(see Tables S8 and S9 in the supplemental materials).
NA Identification Correspondence
Individual Study Results. For the ESM studies, within-strategy correlations representing NA identification correspondence for each regulation strategy ranged from –.44 to .38, with only one effect differing meaningfully from zero. Similarly, for the ESM-Event studies, within-strategy correlations ranged from –.31 to .22 (none reliably non-zero).
Meta-Analyses. In terms of within-strategy correlations, meta-analyses revealed that there was reliable NA identification correspondence for rumination (rmeta = –.09, 95% CI
[–.16 to –.02]), but not for reappraisal or suppression. This finding indicates that although the average NA identification slope for rumination was positive (Table S8), this slope was weaker among individuals with higher global rumination self-reports. We also found evidence of reliable cross-strategy NA identification correspondence: NA identification slopes for state reappraisal were negatively associated with both global suppression (rmeta = –
.15, 95% CI [–.22 to –.08]) and global rumination (rmeta = –.19, 95% CI [–.29 to –.10]); global suppression correlated negatively with NA identification slopes for state rumination
(rmeta = –.08, 95% CI [–.15 to –.01]); and global rumination correlated negatively with NA identification slopes for state suppression (rmeta = –.08, 95% CI [–.15 to –.01]).
PA Identification Correspondence
Individual Study Results. For the ESM studies, within-strategy PA identification correspondence correlations ranged between -.21 and .38, with only one effect differing
EMOTION REGULATION IN EVERYDAY LIFE 23 meaningfully from zero. Similarly, for the ESM-Event studies, within-strategy correlations ranged from –.05 to .28, although none were reliably different from zero.
Meta-Analyses. In terms of within-strategy correlations, results revealed that across studies, there was evidence of PA identification correspondence for suppression (rmeta =.11,
95% CI [.01 to .22]), but not for reappraisal or rumination. This indicates that although the average PA identification slope for state suppression was not significantly different from zero
(Table S9), this slope tended to be more positive among individuals with higher global suppression self-reports.
We also found evidence of cross-strategy PA identification correspondence.
Specifically, global reappraisal was associated with more positive PA identification slopes for state rumination (rmeta =.09, 95% CI [.02 to .16]); and both global suppression (rmeta = .15,
95% CI [.03 to .27]) and global rumination (rmeta = .17, 95% CI [.09 to .26]) were reliably associated with more positive PA identification slopes for state reappraisal.
Implementation Correspondence
Tables 4 and 5 contain within-strategy (shaded) and cross-strategy (unshaded) correlations between global self-reports and within-person NA and PA implementation slopes, respectively. Here, positive correlations indicate that individuals with higher global self-reports tend to have more positive implementation slopes, which reflect the degree to which state use of a regulation strategy predicts changes in subsequent affect. We first report meta-analyses of the average implementation slopes, before reporting NA and PA implementation correspondence findings for the individual studies, focusing on within- strategy correlations. Finally, we report meta-analytic findings, where we discuss both within-strategy and cross-strategy correlations.
EMOTION REGULATION IN EVERYDAY LIFE 24
Table 2 Correlations Among Global Self-Reports and Within-Person Negative Affect (NA) Identification Slopes
NA Identification Slopes (NAt-1 →ERt) "#$% &'%% "'( ),$ ),$ ),$ Global Self-Reports r 95% CI r 95% CI r 95% CI Reappraisal ESM Study 1 (N = 176) .05 -.16 to .26 .18 -.07 to .40 -.08 -.29 to .14 ESM Study 2 (N = 200) .00 -.22 to .20 .20 -.03 to .40 -.04 -.25 to .18 ESM Study 3 (N = 46) -.20 -.77 to .41 .13 -.51 to .69 .01 -.62 to .62 ESM Study 4 (N = 95) .38 .09 to .65 -.15 -.45 to .17 .06 -.28 to .40 ESM-Event Study 1 (N = 101) .22 -.03 to .46 .10 -.16 to .35 .14 -.12 to .40 ESM-Event Study 2 (N = 100) -.18 -.46 to .12 .04 -.30 to .37 -.04 -.35 to .26 ESM-Event Study 3 (N = 112) .09 -.25 to .44 -.14 -.47 to .20 .03 -.31 to .35 Meta-Analysisa (N = 830) .06 -.09 to .21 .06 -.06 to .17 .00 -.07 to .07 Suppression ESM Study 1 (N = 176) -.25 -.44 to -.05 -.04 -.26 to .19 -.04 -.25 to .16 ESM Study 2 (N = 200) -.14 -.33 to .07 .01 -.20 to .22 -.07 -.26 to .14 ESM Study 3 (N = 46) .01 -.61 to .60 -.41 -.85 to .20 .00 -.62 to .60 ESM Study 4 (N = 95) -.03 -.35 to .30 -.20 -.50 to .11 .02 -.31 to .36 ESM-Event Study 1 (N = 101) -.11 -.36 to .14 .14 -.11 to .40 -.13 -.39 to .13 ESM-Event Study 2 (N = 100) -.19 -.48 to .11 -.31 -.61 to .02 -.10 -.41 to .21 ESM-Event Study 3 (N = 112) -.14 -.47 to .20 -.03 -.36 to .32 -.23 -.53 to .10 Meta-Analysisa (N = 830) -.15 -.22 to -.08 -.11 -.24 to .03 -.08 -.15 to -.01 Rumination ESM Study 1 (N = 176) -.21 -.41 to .00 -.16 -.37 to .08 -.01 -.21 to .20 ESM Study 2 (N = 200) -.26 -.45 to -.06 -.04 -.24 to .17 -.04 -.25 to .15 ESM Study 3 (N = 46) .14 -.48 to .71 -.04 -.62 to .58 -.06 -.65 to .56 ESM Study 4 (N = 95) -.37 -.65 to -.08 -.11 -.42 to .20 -.09 -.42 to .24 ESM-Event Study 1 (N = 101) -.05 -.30 to .21 .00 -.25 to .26 -.20 -.43 to .07 ESM-Event Study 2 (N = 100) -.15 -.45 to .14 .03 -.29 to .38 -.17 -.48 to .14 ESM-Event Study 3 (N = 112) -.24 -.57 to .09 -.18 -.51 to .17 -.16 -.47 to .15 Meta-Analysisa (N = 830) -.19 -.29 to -.10 -.08 -.15 to -.01 -.09 -.16 to -.02 Note. Shaded cells contain within-strategy correlations representing identification correspondence for each emotion-regulation strategy; Estimates in bold have 95% CIs that do not include zero. a Meta-analyses based on Fisher z-transformed correlations with frequentist 95% CIs.
Average Implementation Slopes
Meta-analyses of the average within-person implementation slopes revealed that
daily rumination predicted increases in NA and decreases in PA, suppression predicted
increases in NA and reappraisal predicted increases in PA (see Tables S10 and S11).
EMOTION REGULATION IN EVERYDAY LIFE 25
Table 3 Correlations Among Global Self-Reports and Within-Person Positive Affect (PA) Identification Slopes
PA Identification Slopes (PAt-1 →ERt) "#$% &'%" "'( )%$ )%$ )%$ Global Self-Reports r 95% CI r 95% CI r 95% CI Reappraisal ESM Study 1 (N = 176) -.15 -.35 to .06 -.04 -.27 to .17 .08 -.13 to .29 ESM Study 2 (N = 200) -.01 -.26 to .23 -.10 -.36 to .18 .09 -.15 to .31 ESM Study 3 (N = 46) .10 -.51 to .68 -.21 -.75 to .38 .04 -.55 to .64 ESM Study 4 (N = 95) -.21 -.53 to .14 .29 -.04 to .60 .26 -.08 to .57 ESM-Event Study 1 (N = 101) -.03 -.28 to .23 -.01 -.27 to .25 .03 -.23 to .29 ESM-Event Study 2 (N = 100) .06 -.23 to .36 .04 -.31 to .38 -.08 -.40 to .23 ESM-Event Study 3 (N = 112) .11 -.21 to .42 .02 -.27 to .34 .16 -.14 to .45 Meta-Analysisa (N = 830) -.03 -.12 to .06 .00 -.10 to .11 .09 .02 to .16 Suppression ESM Study 1 (N = 176) .28 .09 to .47 .12 -.10 to .33 -.03 -.24 to .17 ESM Study 2 (N = 200) .18 -.07 to .40 -.05 -.30 to .20 .14 -.08 to .34 ESM Study 3 (N = 46) -.03 -.63 to .57 .38 -.21 to .82 .04 -.55 to .63 ESM Study 4 (N = 95) .03 -.30 to .37 .04 -.29 to .38 -.13 -.46 to .18 ESM-Event Study 1 (N = 101) -.10 -.35 to .16 .00 -.27 to .25 .12 -.14 to .37 ESM-Event Study 2 (N = 100) .31 .01 to .58 .15 -.21 to .48 .01 -.31 to .33 ESM-Event Study 3 (N = 112) .25 -.05 to .54 .28 -.01 to .53 .31 .01 to .58 Meta-Analysisa (N = 830) .15 .03 to .27 .11 .01 to .22 .07 -.04 to .18 Rumination ESM Study 1 (N = 176) .30 .10 to .49 .12 -.10 to .34 .03 -.19 to .22 ESM Study 2 (N = 200) .14 -.11 to .36 -.15 -.38 to .10 -.22 -.42 to -.01 ESM Study 3 (N = 46) -.04 -.64 to .55 .12 -.48 to .68 .17 -.43 to .72 ESM Study 4 (N = 95) .26 -.08 to .59 .02 -.32 to .36 -.15 -.47 to .18 ESM-Event Study 1 (N = 101) .02 -.23 to .28 .02 -.24 to .28 .06 -.20 to .31 ESM-Event Study 2 (N = 100) .19 -.10 to .49 .03 -.34 to .38 -.05 -.37 to .27 ESM-Event Study 3 (N = 112) .18 -.12 to .49 .37 .08 to .61 .20 -.09 to .49 Meta-Analysisa (N = 830) .17 .09 to .26 .07 -.06 to .20 -.01 -.13 to .11 Note. Shaded cells contain within-strategy correlations representing identification correspondence for each emotion-regulation strategy; Estimates in bold have 95% CIs that do not include zero. a Meta-analyses based on Fisher z-transformed correlations with frequentist 95% CIs.
NA Implementation Correspondence
Individual Study Results. For the ESM studies, correlations representing NA
implementation correspondence ranged between -.23 and .35, with only one effect reliably
different from zero. Similarly, in the ESM-Event studies, correlations ranged from -.14 to .21
with all 95% CIs including zero, indicating no reliable evidence of NA implementation
correspondence.
EMOTION REGULATION IN EVERYDAY LIFE 26
Table 4 Correlations Among Global Self-Reports and Within-Person Negative Affect (NA) Implementation Slopes
NA Implementation Slopes (ERt →NAt)
&$ &$ &$ '"#$% ''(%" '"() Global Self-Reports r 95% CI r 95% CI r 95% CI Reappraisal ESM Study 1 (N = 176) -.05 -.24 to .14 .06 -.15 to .26 -.17 -.34 to .02 ESM Study 2 (N = 200) -.15 -.35 to .06 -.11 -.30 to .08 -.15 -.34 to .02 ESM Study 3 (N = 46) .06 -.53 to .62 -.11 -.69 to .49 -.16 -.72 to .43 ESM Study 4 (N = 95) -.04 -.38 to .27 -.08 -.37 to .23 .07 -.21 to .34 ESM-Event Study 1 (N = 101) .00 -.27 to .27 .10 -.26 to .41 .04 -.23 to .30 ESM-Event Study 2 (N = 100) -.14 -.43 to .15 .20 -.12 to .49 -.05 -.35 to .24 ESM-Event Study 3 (N = 112) -.04 -.35 to .26 -.16 -.43 to .13 .14 -.14 to .41 Meta-Analysisa (N = 830) -.07 -.14 to -.002 -.01 -.11 to .09 -.04 -.14 to .05 Suppression ESM Study 1 (N = 176) -.18 -.35 to .02 -.23 -.44 to -.03 -.01 -.20 to .17 ESM Study 2 (N = 200) .08 -.13 to .27 .04 -.14 to .23 -.11 -.29 to .07 ESM Study 3 (N = 46) -.09 -.64 to .50 -.04 -.62 to .54 .01 -.58 to .57 ESM Study 4 (N = 95) .34 .00 to .62 -.19 -.47 to .10 -.16 -.41 to .14 ESM-Event Study 1 (N = 101) -.02 -.29 to .25 .02 -.29 to .34 .11 -.16 to .37 ESM-Event Study 2 (N = 100) .18 -.47 to .11 .06 -.27 to .36 .08 -.22 to .35 ESM-Event Study 3 (N = 112) -.01 -.33 to .30 -.11 -.38 to .19 .01 -.27 to .29 Meta-Analysisa (N = 830) -.01 -.14 to .13 -.07 -.17 to .03 -.02 -.09 to .05 Rumination ESM Study 1 (N = 176) -.35 -.52 to -.17 -.18 -.38 to .04 .12 -.07 to .31 ESM Study 2 (N = 200) -.23 -.41 to -.03 -.02 -.22 to .17 .02 -.17 to .21 ESM Study 3 (N = 46) -.27 -.76 to .31 .04 -.54 to .62 .35 -.22 to .81 ESM Study 4 (N = 95) -.03 -.35 to .30 .02 -.29 to .32 .16 -.12 to .44 ESM-Event Study 1 (N = 101) .00 -.27 to .26 .10 -.20 to .41 .01 -.25 to .28 ESM-Event Study 2 (N = 100) -.12 -.39 to .18 .01 -.31 to .31 .21 -.07 to .48 ESM-Event Study 3 (N = 112) .04 -.26 to .36 -.08 -.38 to .21 .03 -.25 to .31 Meta-Analysisa (N = 830) -.14 -.26 to -.02 -.03 -.11 to .04 .10 .03 to .17 Note. Shaded cells contain within-strategy correlations representing implementation correspondence for each emotion-regulation strategy; Estimates in bold have 95% CIs that do not include zero. a Meta-analyses based on Fisher z-transformed correlations with frequentist 95% CIs.
Meta-Analyses. In terms of within-strategy correlations, we found evidence of NA
implementation correspondence for reappraisal (rmeta = –.07, 95% CI [–.14 to –.002]), and
rumination (rmeta = .10, 95% CI [.03 to .17]) but not for suppression. For reappraisal, although
the average NA implementation slope was not significantly different from zero (see Table
S10), individuals scoring higher on global reappraisal tended to have more negative NA
implementation slopes. For rumination, the average NA implementation slope was positive
EMOTION REGULATION IN EVERYDAY LIFE 27
(see Table S10) and was stronger among individuals with higher global rumination scores.
Furthermore, we found evidence of cross-strategy implementation correspondence among
global rumination and state reappraisal (rmeta = –.14, 95% CI [–.26 to –.02]).
Table 5 Correlations Among Global Self-Reports and Within-Person Positive Affect (PA) Implementation Slopes
PA Implementation Slopes (ERt →PAt)
%$ %$ %$ '"#$% ''(%" '"() Global Self-Reports r 95% CI r 95% CI r 95% CI Reappraisal ESM Study 1 (N = 176) .05 -.14 to .23 .05 -.16 to .26 .12 -.07 to .30 ESM Study 2 (N = 200) -.04 -.29 to .23 -.04 -.27 to .19 .10 -.11 to .32 ESM Study 3 (N = 46) -.14 -.70 to .48 .02 -.57 to .61 .10 -.51 to .68 ESM Study 4 (N = 95) .19 -.18 to .51 .30 -.04 to .61 -.16 -.44 to .13 ESM-Event Study 1 (N = 101) -.18 -.43 to .08 -.08 -.40 to .26 -.01 -.28 to .27 ESM-Event Study 2 (N = 100) .11 -.21 to .40 -.05 -.34 to .26 .09 -.21 to .38 ESM-Event Study 3 (N = 112) -.01 -.31 to .27 .15 -.12 to .43 .02 -.25 to .28 Meta-Analysis (N = 830) .01 -.08 to .09 .05 -.05 to .15 .05 -.02 to .12 Suppression ESM Study 1 (N = 176) .28 .10 to .44 .27 .06 to .46 .13 -.06 to .31 ESM Study 2 (N = 200) .04 -.22 to .29 -.05 -.27 to .17 .05 -.15 to .26 ESM Study 3 (N = 46) -.08 -.64 to .54 .17 -.41 to .72 -.20 -.74 to .39 ESM Study 4 (N = 95) -.11 -.47 to .26 -.05 -.38 to .28 .12 -.18 to .40 ESM-Event Study 1 (N = 101) .06 -.21 to .32 .06 -.24 to .36 -.05 -.33 to .23 ESM-Event Study 2 (N = 100) .19 -.12 to .47 .06 -.25 to .34 .14 -.15 to .41 ESM-Event Study 3 (N = 112) .13 -.19 to .42 .07 -.23 to .34 -.24 -.48 to .03 Meta-Analysis (N = 830) .09 -.02 to .20 .07 -.03 to .17 .01 -.11 to .12 Rumination ESM Study 1 (N = 176) .49 .34 to .64 .17 -.04 to .38 -.04 -.23 to .15 ESM Study 2 (N = 200) .11 -.15 to .35 -.07 -.29 to .16 .03 -.18 to .25 ESM Study 3 (N = 46) .15 -.47 to .70 -.28 -.75 to .33 .10 -.47 to .68 ESM Study 4 (N = 95) .03 -.32 to .41 -.03 -.38 to .32 -.11 -.40 to .19 ESM-Event Study 1 (N = 101) -.04 -.30 to .22 -.09 -.39 to .20 -.03 -.30 to .24 ESM-Event Study 2 (N = 100) .18 -.13 to .47 .08 -.22 to .37 -.15 -.43 to .14 ESM-Event Study 3 (N = 112) .00 -.29 to .29 .09 -.20 to .38 -.05 -.36 to .18 Meta-Analysis (N = 830) .14 -.01 to .30 .00 -.10 to .10 -.04 -.11 to .03 Note. Shaded cells contain within-strategy correlations representing implementation correspondence for each emotion-regulation strategy; Estimates in bold have 95% CIs that do not include zero. a Meta-analyses based on Fisher z-transformed correlations with frequentist 95% CIs.
PA Implementation Correspondence
Individual Study Results. In the ESM studies, correlations representing PA
implementation correspondence ranged between -.14 and .27, only one of which was
EMOTION REGULATION IN EVERYDAY LIFE 28 meaningfully different from zero. In the ESM-Event studies, correlations ranged from -.18 to
.11 with all 95% CIs including zero, indicating no evidence of reliable PA implementation correspondence.
Meta-Analyses. Meta-analyses across all ESM and ESM-Event studies revealed no reliable evidence of PA implementation correspondence either within or across emotion- regulation strategies.
Unique Effects of Selection, Identification, and Implementation on Global Self-Reports
To test whether our main correlational findings were robust when controlling for potential overlap among the three components of daily emotion regulation, we repeated our main analyses using multiple regression. Global scores on each strategy were simultaneously regressed onto the between-person component of daily self-reported strategy use
(representing selection) as well as identification and implementation slopes for each strategy.
This allowed us to estimate unique associations between each component of daily emotion- regulation strategy use and global self-reports of the same regulation strategy. Results of these models are reported in Tables S12 and S13 in the supplemental materials. To reduce multicollinearity among predictors of global emotion regulation and make these analyses tractable, we ran separate models for NA (see Table S12) and PA (see Table S13), and no cross-strategy effects were estimated.
These analyses largely mirror those we report above in the correlational analyses, revealing that global reappraisal was not reliably predicted by selection, identification, or implementation of state reappraisal, for either NA or PA. In contrast, global suppression and rumination scores were uniquely predicted by their respective between-person daily use
(representing selection correspondence). Furthermore, global rumination was also uniquely negatively predicted by NA identification and positively predicted by NA implementation
EMOTION REGULATION IN EVERYDAY LIFE 29
slopes. In addition, global suppression was also uniquely positively predicted by PA
identification slopes.
Discussion
Thousands of studies have used global self-report questionnaires to assess habitual
emotion-regulation strategy selection. Yet, we know relatively little about what such global
measures assess. To begin understanding this, we investigated how people’s global self-
reports of emotion regulation correspond with the daily dynamics of emotion regulation.
Drawing on the distinct stages of emotion regulation proposed in Gross’s (2015) extended
process model and other recent theoretical accounts (e.g., Webb et al., 2012b), we mapped
global emotion regulation measures onto operationalizations of daily identification, selection,
and implementation of emotion-regulation strategies across nine diary and ESM studies.
Table 6 provides a high-level summary of our findings across strategies and stages
of the emotion regulation process, considering both the correlational analyses reported above
(Tables 1 to 5) and simultaneous regressions reported in the supplementary materials (Tables
S12 to S13) that controlled for shared variance between selection, identification, and
implementation of each strategy in daily life. In the following sections, we discuss the
findings for each emotion regulation stage.
Table 6 Summary of Results Across Studies, Strategies, and Stages of Emotion Regulation Theory. Emotion-regulation strategy Emotion Regulation Stage Reappraisal Suppression Rumination Selection ? (+) ü (+) ü (+) Identification Negative affect û û ü (–) Positive affect û ü (+) û Implementation Negative affect ? (–) û ü (+) Positive affect û û û Note. ü consistently significant associations; û consistently non-significant associations; ? inconsistent results across analyses; + positive association; – negative association.
EMOTION REGULATION IN EVERYDAY LIFE 30
Selection Correspondence
Our meta-analyses revealed weak-to-moderate positive correlations (.14 ≤ rs ≤ .40) between global self-reports and the stable between-person component of daily self-reports, representing trait selection of emotion-regulation strategies (see Table 1). These relationships are weaker than what is typically observed for broad personality dimensions, such as the Big
Five domains (e.g., Augustine & Larsen, 2012; Finnigan & Vazire, 2018; Fleeson &
Gallagher, 2009). Rather, our findings are more similar to the degree of correspondence between global and momentary self-reports of narrower constructs, such as dispositional anxiety and anger (Edmondson et al., 2013) or the Big Five aspects (Rauthmann et al., 2018).
We also found that selection correspondence differed across strategies (see Table 1): we found stronger evidence of selection correspondence for rumination and suppression than for reappraisal. Indeed, the weak meta-analytic correlation between global and mean-daily reappraisal selection (r = .14) suggests that the ERQ reappraisal scale should not be interpreted directly as an index of trait reappraisal selection. This may be problematic given the widespread interpretation of this scale as a measure of habitual reappraisal. The evidence for suppression selection correspondence was comparatively stronger, suggesting that the
ERQ suppression subscale may better index regular use of this strategy than its reappraisal- subscale counterpart.
In contrast, the estimated correlation between global and momentary self-reports was more than twice as large for rumination than for reappraisal, suggesting that people may be substantially more accurate in reporting their habitual tendencies to ruminate than their habitual use of reappraisal in daily life. We offer two possible reasons for the stronger selection correspondence observed for rumination relative to reappraisal.
First, unlike reappraisal, rumination is inherently perseverative, temporally extended, and intrusive (Kircanski et al., 2015; Nolen-Hoeksema et al., 2008), which may
EMOTION REGULATION IN EVERYDAY LIFE 31 make daily use of this strategy more memorable. This should lead to stronger memory encoding for episodes of rumination, which people may draw on when providing global self- reports (Schwarz, 2012), resulting in greater selection correspondence for rumination.
Second, our analyses suggest that rumination may be more strongly tied with affective experience in daily life than reappraisal. Specifically, relative to reappraisal, momentary rumination was more strongly (i) predicted by affect intensity at the previous occasion (see
Tables S8 and S9 in supplemental materials), and (ii) predictive of future affect intensity (see
Tables S10 and S11 in supplemental materials). Given that emotional experiences tend to be weighted heavily in human experience (e.g., Kensinger, 2009; Rozin & Royzman, 2001), the stronger association between affect and rumination may make it easier for people to recall episodes of rumination (relative to reappraisal) when aggregating across many occasions to form a global judgment. This is an example of what Wrzus and Roberts (2017) refer to as an associative mechanism linking repeated short-term state dynamics with long-term trait development.
Put otherwise, people may be more likely to engage in rumination than reappraisal when experiencing unpleasant feelings, and rumination is more likely than reappraisal to influence people’s subsequent affective experience. This type of mutually reinforcing
“downward spiral” may be memorable and therefore more likely to be integrated into people’s self-beliefs, leading to fairly strong correspondence between global and momentary self-reports of rumination. In contrast, reappraisal may be less uniformly triggered by negative emotions (Suri et al., 2015) and its affective consequences may be more context- dependent (Ford & Troy, 2019; Haines et al., 2016), making it more difficult to accurately report habitual reappraisal selection on global self-report measures.
A second explanation for the relatively stronger selection correspondence observed for rumination is that global self-reports of rumination may be a proxy for general distress.
EMOTION REGULATION IN EVERYDAY LIFE 32
This may partly account for our finding that global rumination scores also correlated reliably with the between-person components of daily reappraisal and suppression (see Table 1). If global rumination is a proxy for general distress (or equivalently, a tendency to experience greater NA and lower PA in daily life), it may be associated with higher regulatory effort across multiple strategies. Supporting this view, Aldao et al.’s (2010) meta-analysis revealed that global rumination scores correlated more strongly with psychopathology symptoms than global self-reports of other emotion-regulation strategies.
Identification Correspondence
Our meta-analyses for NA identification correspondence revealed one consistent finding across studies and analytic approaches: global rumination scores correlated negatively with NA identification slopes for state rumination. To understand this finding, recall that the average NA identification slope for state rumination was positive (see Table
S8), indicating that state rumination tended to increase following more intense NA. Thus, for people with high global rumination scores, state rumination was less contingent upon preceding NA levels than among those with low global rumination scores. Similarly, global rumination was negatively related to NA identification slopes for state reappraisal and suppression (see Table 2), indicating that state use of all three regulation strategies was less contingent on previous NA among people with high global self-reported rumination. Taken together with our findings for selection correspondence, these results suggest that global rumination is associated with greater, but also less context-sensitive, use of multiple emotion- regulation strategies in daily life. Given that global self-reported rumination is strongly associated with vulnerability to depression (Nolen-Hoeksema et al., 2008), this aligns with the literature on emotion context insensitivity (Rottenberg & Hindash, 2015).
One possible explanation for these findings is that people who report high levels of global rumination may overrepresent the need to regulate unpleasant affect. Sheppes et al.
EMOTION REGULATION IN EVERYDAY LIFE 33
(2015) proposed this as a possible mechanism underlying difficulties in the identification stage of emotion regulation, which may trigger persistent and ultimately maladaptive regulatory efforts. Put otherwise, individuals who report high levels of global rumination may tend to deploy multiple emotion-regulation strategies inflexibly across a range of NA intensities. Although the most widely used measure of global rumination (i.e., the RRS) instructs participants to indicate how much they ruminate specifically when feeling “down, sad, or depressed”, it instead appears to measure a tendency to use multiple regulation strategies (including rumination) relatively independent of preceding levels of unpleasant affect.
In terms of PA identification correspondence, we found that global suppression was associated with more positive PA identification slopes for state suppression. Thus, although the average PA identification slope for state suppression was not significant (see Table S9), high global suppressors were more likely to use suppression in daily life when experiencing above-average PA. Here, our findings deviate from previous research examining how global measures of suppression relate to affect-dependent use of state suppression in the lab and in daily life (Peters et al., 2020). For instance, whereas Peters and colleagues reported that ERQ suppression scores were associated with a greater tendency to deploy situational suppression in response to negative emotions, we did not find reliable evidence of NA identification correspondence. This may be because we assessed state suppression in everyday life, where negative emotions may be less frequent and intense, compared to the contexts of negative social interactions and romantic conflicts studied by Peters et al. (2020). Although Peters et al. (2020) did not assess positive emotions, the current findings align with their speculation that the ERQ suppression subscale may correlate with greater state use of suppression in response to PA. This finding is also broadly consistent with previous research linking global
EMOTION REGULATION IN EVERYDAY LIFE 34 self-reported suppression with a tendency to inhibit the expression of positive emotions
(Kashdan & Breen, 2008).
Implementation Correspondence
Our meta-analyses testing implementation correspondence revealed that global reappraisal correlated negatively with NA implementation slopes for state reappraisal.
Although state reappraisal did not predict decreases in NA, on average (see meta-analysis in
Table S10), people scoring higher on the ERQ reappraisal subscale tended to report larger
NA decreases following their use of reappraisal in daily life. This suggests, in line with some past work (Ford et al., 2017; McRae et al., 2012), that the ERQ reappraisal subscale may measure how effectively reappraisal is implemented as well as indexing its habitual selection.
Thus, although reappraisal may be relatively difficult to implement, and therefore had no overall influence on unpleasant feelings in daily life, individuals who use reappraisal more often may become more proficient at implementing it (cf. Denny & Ochsner, 2014).
However, tempering the above conclusions, this finding was not robust to controlling for individual differences in identification and selection of reappraisal in daily life
(see multiple regression results in Table S12). Furthermore, global rumination scores were also associated with more negative NA implementation slopes for state reappraisal, indicating that people who reported being high in habitual rumination also showed larger decreases in
NA following state reappraisal, an indicator of more effective implementation of reappraisal in daily life. One possibility is that processes common to rumination and reappraisal may play a role here, but given the dearth of studies on such shared mechanisms this requires further investigation.
Finally, global rumination scores correlated positively with NA implementation slopes for state rumination. Given that, on average, state rumination predicted increases in
NA across time (see Table S10), this indicates that state rumination is more harmful (in terms
EMOTION REGULATION IN EVERYDAY LIFE 35 of increasing NA) among individuals scoring higher on global rumination, in line with previous findings (Gerin et al., 2006; Johnson et al., 2012). When considered together with our findings for NA identification, this suggests that in addition to measuring a greater tendency to select rumination as a regulation strategy in daily life, global rumination scales appear to correlate with lower sensitivity to negative affect cues, in terms of identifying the need to ruminate, and a stronger impact of implementing state rumination on subsequent NA.
Thus, high global ruminators appear vulnerable to mis-regulation cycles, in which they engage in state rumination more often, regardless of how unpleasant they are feeling, and with more detrimental effects on their subsequent affective experience.
Theoretical and Methodological Implications
Recent theoretical advances in emotion regulation have sought to understand emotion regulation as a dynamic, multi-stage process (Gross, 2015) with broad implications for understanding what constitutes (un)healthy emotion regulation (e.g., Aldao et al., 2015;
Sheppes, 2020). Relatedly, several researchers have called for the use of daily life methods to obtain more dynamic, context-specific assessments of emotion regulation (e.g., Aldao, 2013;
Colombo et al., 2020; McMahon & Naragon-Gainey, 2020). While we agree that daily life methods are an essential addition to our research toolbox, we recognize that global self-report questionnaires remain important measurement instruments for emotion regulation researchers. Thus, our aim in the current study was to model how these two types of measures relate to one another, to better understand how to deploy them to accurately measure the stages of emotion regulation.
Our results caution against a straightforward interpretation of global measures as representing individual differences in strategy selection, for two reasons. First, our meta- analytic findings indicate that global self-report measures of a given emotion-regulation strategy do not align strongly or specifically with trait selection of that strategy in daily life.
EMOTION REGULATION IN EVERYDAY LIFE 36
In particular, the ERQ reappraisal subscale appears to correspond only weakly with the tendency to select reappraisal in daily life. In contrast, global self-reports of suppression and rumination showed stronger correspondence with average daily strategy selection. However, these associations were not strategy-specific: global rumination scores correlated reliably with average daily reappraisal and suppression, and global suppression correlated with daily rumination, undermining the discriminant validity of these global measures.
Second, our meta-analyses across studies showed that global self-report measures of emotion regulation were reliably associated with individual differences in the affective antecedents and consequences of regulation strategies in daily life, which we argue are driven by identification and implementation processes. Specifically, global suppression and rumination scores were consistently correlated with individual differences in daily identification and implementation slopes. Furthermore, global suppression scores correlated with identification slopes for PA but not NA, whereas global rumination scores were associated with identification and implementation for NA but not PA. Overall, our findings suggest that global self-report measures capture other dynamic aspects of the regulation process beyond habitual strategy selection.
Beyond emotion regulation, our findings for selection correspondence have implications for theorizing on the alignment between global and momentary self-reports in general. Specifically, our results are more consistent with Robinson and Clore’s (2002) accessibility model of emotional self-report than with Fleeson and colleagues’ whole trait theory. Whereas whole trait theory explains the relatively strong correspondence between global and daily measures of the Big Five (Fleeson & Jayawickreme, 2015), it does not appear to generalize as readily to narrower dispositions, such as trait selection of emotion- regulation strategies. This may be because the Big Five represent broad personality dimensions—each subsuming several affective, behavioural, and cognitive tendencies
EMOTION REGULATION IN EVERYDAY LIFE 37
(DeYoung, 2015)—whereas emotion-regulation strategies are specific cognitive or behavioral actions. Thus, there may be stronger selection correspondence for higher-order dimensions of emotion-regulation strategies (e.g., avoidance, engagement) than for specific strategies (cf. McMahon & Naragon-Gainey, 2020). However, our findings also suggest that, at least in relation to emotion regulation, global self-report measures may not only index the location (i.e., central tendency) of a person’s density distribution of states (Fleeson &
Jayawickreme, 2015), but may also reflect individual differences in dynamics, such as the temporal contingencies between state emotion regulation and state affect (cf. Geukes et al.,
2018).
Limitations and Future Directions
We note several limitations of the current study and suggest possible directions for future research. First and foremost, our operationalizations of the three stages of emotion regulation in daily life are imperfect and incomplete. In particular, for parsimony we considered the within-person slopes indexing the affective antecedents and consequences of each emotion-regulation strategy as operationalizations of identification and implementation, respectively. However, these slopes are unlikely to capture the entirety of the identification and implementation stages, and they may also be influenced by processes occurring during strategy selection. For example, our operationalization of identification did not directly represent the activation of a goal to regulate–a key element of the identification stage (Gross,
2015). Similarly, because we operationalized identification in a strategy-specific manner, we may have also captured strategy selection to a certain extent. Future research should explore alternate operationalizations of identification and implementation in daily life, such as self- reported regulation goals (identification) and regulation success (implementation). The addition of such measures would provide an opportunity to validate our proposed operationalizations of identification and implementation.
EMOTION REGULATION IN EVERYDAY LIFE 38
Second, we focused on three strategies among many that people routinely use to regulate emotions in daily life (Heiy & Cheavens, 2014; Parkinson & Totterdell, 1999). We selected these three strategies because they are among the most widely studied in the emotion regulation literature (see e.g., Naragon-Gainey et al., 2017; Webb et al., 2012a). Thus, we believe that despite our limited focus, the current investigation has relevance to a large proportion of emotion regulation research. Nevertheless, future studies should investigate correspondence between global and daily self-reports of other emotion-regulation strategies.
Two recent examples of such research are studies by Lavender et al. (2017) and Medland et al. (2020) examining correspondence between global and daily versions of the Difficulties in
Emotion Regulation Scale (Gratz & Roemer, 2004) and the Regulating Emotion Systems
Survey (De France & Hollenstein, 2017), respectively. Although these studies make important contributions, they focus on the (dys)regulation of negative emotions. More research is needed on correspondence between global and daily measures of positive emotion-regulation strategies, which are generally under-researched (Quoidbach et al., 2010).
Third, the current study cannot speak to the validity of daily self-reports. Although daily self-reports reduce retrospective biases by assessing emotion-regulation strategies closer to the time and context in which they are used, they are not immune from other sources of bias (Conner & Barrett, 2012; Finnigan & Vazire, 2018; Schwarz, 2012). For instance, daily self-reports may be biased because people have imperfect insight into their behavior
(Sun & Vazire, 2019) or because they are motivated to deny socially undesirable behavior
(Gosling et al., 1998). Further research is needed to establish the validity of daily measures of emotion regulation. However, recent research examining the correspondence between daily self-reported feelings and observed verbal (Sun & Vazire, 2019) and written (Kross et al.,
2019) expression of emotions highlights the challenges in finding a suitable validation criterion for daily self-reports. Given the limitations inherent in all self-report methods
EMOTION REGULATION IN EVERYDAY LIFE 39
(Paulhus & Vazire, 2007), we submit that neither global nor daily self-reports should be considered “gold-standard” measures of emotion-regulation strategy use in daily life.
Finally, the current study leaves open the question of whether daily self-reports of emotion regulation are more or less predictive of important outcomes than global self-reports.
Even if global emotion regulation scales do not correspond strongly with people’s habitual selection of emotion-regulation strategies, they may nevertheless predict relevant outcomes more strongly than daily self-reports (e.g., Finnigan & Vazire, 2018). Many studies have established the predictive validity of global self-report measures of emotion-regulation strategies for predicting well-being and psychological functioning (Aldao et al., 2010). The challenge for future research on emotion regulation is to develop a better understanding of which aspects of daily emotion regulation processes are captured by global self-report scales and to investigate whether the dynamics of emotion regulation captured with daily self- reports have incremental predictive validity over and above global measures.
Concluding Remarks
Research on emotion regulation has grown exponentially over the past few decades and much of this work has relied on global self-report measures to assess individual differences in emotion regulation. In this study, we mapped global self-report measures to operationalizations of identification, selection, and implementation of emotion-regulation strategies in daily life. Overall our findings suggest that, with the exception of reappraisal, global self-report measures correspond quite well with trait strategy selection in daily life, but not necessarily in a strategy- or process-specific manner. Thus, global self-reports of one emotion-regulation strategy may index individual differences in selection of other strategies, as well as processes that may be better conceptualized as occurring during the identification and implementation stages of regulation.
EMOTION REGULATION IN EVERYDAY LIFE 40
References
Aldao, A. (2013). The future of emotion regulation research: Capturing context. Perspectives on Psychological Science, 8(2), 155–172. https://doi.org/10.1177/1745691612459518
Aldao, A., Nolen-Hoeksema, S., & Schweizer, S. (2010). Emotion-regulation strategies across psychopathology: A meta-analytic review. Clinical Psychology Review, 30(2),
217–237. https://doi.org/10.1016/j.cpr.2009.11.004
Aldao, A., Sheppes, G., & Gross, J. J. (2015). Emotion regulation flexibility. Cognitive Therapy and Research, 39(3), 263–278. https://doi.org/10.1007/s10608-014-9662-4
Asparouhov, T., & Muthén, B. (2019). Latent variable centering of predictors and mediators
in multilevel and time-series models. Structural Equation Modeling, 26, 119-142. https://doi.org/10.1080/10705511.2018.1511375
Augustine, A. A., & Larsen, R. J. (2012). Is a trait really the mean of states? Similarities and differences between traditional and aggregate assessments of personality. Journal of
Individual Differences, 33(3), 131–137. https://doi.org/10.1027/1614-0001/a000083
Blanke, E. S., Brose, A., Kalokerinos, E. K., Erbas, Y., Riediger, M., & Kuppens, P. (2019). Mix it to fix it: Emotion regulation variability in daily life. Emotion, 20(3), 473–485.
https://doi.org/10.1037/emo0000566 Brans, K., Koval, P., Verduyn, P., Lim, Y. L., & Kuppens, P. (2013). The regulation of
negative and positive affect in daily life. Emotion, 13(5), 926–939.
https://doi.org/10.1037/a0032400 Brockman, R., Ciarrochi, J., Parker, P., & Kashdan, T. (2017). Emotion regulation strategies
in daily life: Mindfulness, cognitive reappraisal and emotion suppression. Cognitive
Behaviour Therapy, 46(2), 91–113. https://doi.org/10.1080/16506073.2016.1218926 Colombo, D., Fernández-Álvarez, J., Suso-Ribera, C., Cipresso, P., Valev, H., Leufkens, T.,
Sas, C., Garcia-Palacios, A., Riva, G., & Botella, C. (2020). The need for change: Understanding emotion regulation antecedents and consequences using ecological
momentary assessment. Emotion, 20(1), 30–36. https://doi.org/10.1037/emo0000671
EMOTION REGULATION IN EVERYDAY LIFE 41
Conner, T. S., & Barrett, L. F. (2012). Trends in ambulatory self-report: The role of
momentary experience in psychosomatic medicine. Psychosomatic Medicine, 74(4), 327–337. https://doi.org/10.1097/PSY.0b013e3182546f18.
De France, K., & Hollenstein, T. (2017). Assessing emotion regulation repertoires: The Regulation of Emotion Systems Survey. Personality and Individual Differences, 119,
204–215. https://doi.org/10.1016/j.paid.2017.07.018
Denny, B. T., & Ochsner, K. N. (2014). Behavioral effects of longitudinal training in cognitive reappraisal. Emotion, 14(2), 425-433. https://doi.org/10.1037/a0035276
DeYoung, C. G. (2015). Cybernetic Big Five Theory. Journal of Research in Personality, 56,
33–58. https://doi.org/10.1016/j.jrp.2014.07.004 Edmondson, D., Shaffer, J. A., Chaplin, W. F., Burg, M. M., Stone, A. A., & Schwartz, J. E.
(2013). Trait anxiety and trait anger measured by ecological momentary assessment and their correspondence with traditional trait questionnaires. Journal of Research in
Personality, 47(6), 843-852. https://doi.org/10.1016/j.jrp.2013.08.005
English, T., Lee, I. A., John, O. P., & Gross, J. J. (2017). Emotion regulation strategy selection in daily life: The role of social context and goals. Motivation and Emotion,
41(2), 230–242. https://doi.org/10.1007/s11031-016-9597-z Feldman, J. L., & Freitas, A. L. (2021). The generality of effects of emotional experience on
emotion-regulation choice. Emotion, 21(1), 211–219.
https://doi.org/10.1037/emo0000611 Finnigan, K. M., & Vazire, S. (2018). The incremental validity of average state self-reports
over global self-reports of personality. Journal of Personality and Social Psychology,
115(2), 321–337. https://doi.org/10.1037/pspp0000136 Fleeson, W., & Gallagher, P. (2009). The implications of Big Five standing for the
distribution of trait manifestation in behavior: Fifteen experience-sampling studies and a meta-analysis. Journal of Personality and Social Psychology, 97(6), 1097–
1114. https://doi.org/10.1037/a0016786
Fleeson, W., & Jayawickreme, E. (2015), Whole Trait Theory. Journal of Research in Personality, 56, 82-92. https://doi.org/10.1016/j.jrp.2014.10.009
EMOTION REGULATION IN EVERYDAY LIFE 42
Ford, B. Q., Karnilowicz, H. R., & Mauss, I. B. (2017). Understanding reappraisal as a
multicomponent process: The psychological health benefits of attempting to use reappraisal depend on reappraisal success. Emotion, 17(6), 905–911.
https://doi.org/10.1037/emo0000310 Ford, B. Q., & Troy, A. S. (2019). Reappraisal reconsidered: A closer look at the costs of an
acclaimed emotion-regulation strategy. Current Directions in Psychological Science,
28(2), 195–203. https://doi.org/10.1177/0963721419827526 Garnefski, N., Kraaij, V., & Spinhoven, P. (2001). Negative life events, cognitive emotion
regulation and emotional problems. Personality and Individual Differences, 30(8),
1311–1327. https://doi.org/10.1016/S0191-8869(00)00113-6 Gerin, W., Davidson, K. W., Christenfeld, N. J., Goyal, T., & Schwartz, J. E. (2006). The
role of angry rumination and distraction in blood pressure recovery from emotional arousal. Psychosomatic Medicine, 68, 64-72. https://doi.org/0.1097/01.psy.000019574
Geukes, K., van Zalk, M., & Back, M. D. (2018). Understanding personality development:
An integrative state process model. International Journal of Behavioral Development, 42(1), 43–51. https://doi.org/10.1177/0165025416677847
Gosling, S. D., John, O. P., Craik, K. H., & Robins, R. W. (1998). Do people know how they behave? Self-reported act frequencies compared with on-line codings by observers.
Journal of Personality and Social Psychology, 74(5), 1337–1349.
https://doi.org/10.1037/0022-3514.74.5.1337 Gratz, K. L., & Roemer, L. (2004). Multidimensional assessment of emotion regulation and
dysregulation: Development, factor structure, and initial validation of the difficulties
in emotion regulation scale. Journal of Psychopathology and Behavioral Assessment, 26, 41–54. https://doi.org/10.1023/B:JOBA.0000007455.08539.94
Gross, J. J. (1998). The emerging field of emotion regulation: An integrative review. Review of General Psychology, 2(3), 271–299. https://doi.org/10.1037/1089-2680.2.3.271
Gross, J. J. (2015). Emotion regulation: Current status and future prospects. Psychological
Inquiry, 26(1), 1–26. https://doi.org/10.1080/1047840X.2014.940781
EMOTION REGULATION IN EVERYDAY LIFE 43
Gross, J. J., & John, O. P. (2003). Individual differences in two emotion regulation processes:
Implications for affect, relationships, and well-being. Journal of Personality and Social Psychology, 85(2), 348–362. https://doi.org/10.1037/0022-3514.85.2.348
Gross, J. J., Uusberg, H., & Uusberg, A. (2019). Mental illness and well‐being: an affect regulation perspective. World Psychiatry, 18(2), 130–139.
https://doi.org/10.1002/wps.20618
Haines, S. J., Gleeson, J., Kuppens, P., Hollenstein, T., Ciarrochi, J., Labuschagne, I., Grace, C., & Koval, P. (2016). The wisdom to know the difference: Strategy-situation fit in
emotion regulation in daily life is associated with well-being. Psychological Science,
27(12), 1651–1659. https://doi.org/10.1177/0956797616669086 Hamaker, E. L., Nesselroade, J. R., & Molenaar, P. C. M. (2007). The integrated trait–state
model. Journal of Research in Personality, 41(2), 295–315. https://doi.org/10.1016/j.jrp.2006.04.003
Hamaker, E. L., Schuurman, N. K., & Zijlmans, E. A. O. (2017). Using a few snapshots to
distinguish mountains from waves: Weak factorial invariance in the context of trait- state research. Multivariate Behavioral Research, 52(1), 47–60.
https://doi.org/10.1080/00273171.2016.1251299 Heiy, J. E., & Cheavens, J. S. (2014). Back to basics: A naturalistic assessment of the
experience and regulation of emotion. Emotion, 14(5), 878–891.
https://doi.org/10.1037/a0037231 John, O. P., & Eng, J. (2014). Three approaches to individual differences in affect regulation:
Conceptualizations, measures, and findings. In J. J. Gross (Ed.), Handbook of emotion
regulation (pp. 321-345). Guilford Press. Johnson, J. A., Lavoie, K. L., Bacon, S. L., Carlson, L. E., & Campbell, T. S. (2012). The
effect of trait rumination on adaptation to repeated stress. Psychosomatic Medicine, 74(3), 258-262. https://doi.org/10.1097/PSY.0b013e31824c3ef2
Kalokerinos, E. K., Tamir, M., & Kuppens, P. (2017). Instrumental motives in negative emotion
regulation in daily life: Frequency, consistency, and predictors. Emotion, 17(4), 648-657. https://doi.org/10.1037/emo0000269
EMOTION REGULATION IN EVERYDAY LIFE 44
Kanske, P., Schönfelder, S., Forneck, J., & Wessa, M. (2015). Impaired regulation of emotion:
Neural correlates of reappraisal and distraction in bipolar disorder and unaffected relatives. Translational Psychiatry, 5, e497. https://doi.org/10.1038/tp.2014.137
Kashdan, T. B., & Breen, W. E. (2008). Social anxiety and positive emotions: A prospective examination of a self-regulatory model with tendencies to suppress or express emotions as a
moderating variable. Behavior Therapy, 39(1), 1-12.
https://doi.org/10.1016/j.beth.2007.02.003 Kensinger, E. A. (2009). Remembering the details: Effects of emotion. Emotion review, 1(2), 99-
113. https://doi.org/10.1177/1754073908100432
Kircanski, K., Thompson, R. J., Sorenson, J., Sherdell, L., & Gotlib, I. H. (2015). Rumination and worry in daily life: Examining the naturalistic validity of theoretical constructs. Clinical
Psychological Science, 3(6), 926–939. https://doi.org/10.1177/2167702614566603 Koole, S. L. (2009). The psychology of emotion regulation: An integrative review. Cognition
and Emotion, 23(1), 4–41. https://doi.org/10.1080/02699930802619031
Kross, E., Verduyn, P., Boyer, M., Drake, B., Gainsburg, I., Vickers, B., Ybarra, O., & Jonides, J. (2019). Does counting emotion words on online social networks provide a
window into people’s subjective experience of emotion? A case study on Facebook. Emotion, 19(1), 97–107. https://doi.org/10.1037/emo0000416
Lavender, J. M., Tull, M. T., DiLillo, D., Messman-Moore, T., & Gratz, K. L. (2017).
Development and validation of a state-based measure of emotion dysregulation. Assessment, 24(2), 197–209. https://doi.org/10.1177/1073191115601218
Lennarz, H. K., Hollenstein, T., Lichtwarck-Aschoff, A., Kuntsche, E., & Granic, I. (2019).
Emotion regulation in action: Use, selection, and success of emotion regulation in adolescents’ daily lives. International Journal of Behavioral Development, 43(1), 1–
11. https://doi.org/10.1177/0165025418755540 Livingstone, K. M., & Isaacowitz, D. M. (2019). Age and emotion regulation in daily life:
Frequency, strategies, tactics, and effectiveness. Emotion. Advance online
publication. https://doi.org/10.1037/emo0000672
EMOTION REGULATION IN EVERYDAY LIFE 45
McMahon, T. P., & Naragon-Gainey, K. (2020). Ecological validity of trait emotion
regulation strategy measures. Psychological Assessment, 32(8), 796–802. https://doi.org/10.1037/pas0000827
McRae, K., Jacobs, S. E., Ray, R. D., John, O. P., & Gross, J. J. (2012). Individual differences in reappraisal ability: Links to reappraisal frequency, well-being, and
cognitive control. Journal of Research in Personality, 46(1), 2–7.
https://doi.org/10.1016/j.jrp.2011.10.003 Medland, H., De France, K., Hollenstein, T., Mussoff, D., & Koval, P. (2020). Regulating
emotion Systems in everyday life: Reliability and validity of the RESS-EMA scale.
European Journal of Psychological Assessment, 36(3), 437–446. https://doi.org/10.1027/1015-5759/a000595
Morawetz, C., Bode, S., Derntl, B., & Heekeren, H. R. (2017). The effect of strategies, goals and stimulus material on the neural mechanisms of emotion regulation: A meta-
analysis of fMRI studies. Neuroscience and Biobehavioral Reviews, 72, 111-128.
https://doi.org/10.1016/j.neubiorev.2016.11.014 Muthén, L.K. and Muthén, B.O. (1998-2017). Mplus User’s Guide. Eighth Edition. Muthén
& Muthén. Naragon-Gainey, K., McMahon, T. P., & Chacko, T. P. (2017). The structure of common
emotion regulation strategies: A meta-analytic examination. Psychological Bulletin,
143(4), 384–427. https://doi.org/10.1037/bul0000093 Nezlek, J. B., & Kuppens, P. (2008). Regulating positive and negative emotions in daily life.
Journal of Personality, 76(3), 561–580. https://doi.org/10.1111/j.1467-
6494.2008.00496.x Nolen-Hoeksema, S., & Morrow, J. (1991). A prospective study of depression and
posttraumatic stress symptoms after a natural disaster: The 1989 Loma Prieta earthquake. Journal of Personality and Social Psychology, 61(1), 115–121.
https://doi.org/10.1037/0022-3514.61.1.115
EMOTION REGULATION IN EVERYDAY LIFE 46
Nolen-Hoeksema, S., Wisco, B. E., & Lyubomirsky, S. (2008). Rethinking rumination.
Perspectives on Psychological Science, 3(5), 400–424. https://doi.org/10.1111/j.1745- 6924.2008.00088.x
Parkinson, B., & Totterdell, P. (1999). Classifying affect-regulation strategies. Cognition and Emotion, 13(3), 277–303. https://doi.org/10.1080/026999399379285
Pasyugina, I., Koval, P., De Leersnyder, I., Mesquita, B., & Kuppens, P. (2015).
Distinguishing between level and impact of rumination as predictors of depressive symptoms: An experience sampling study. Cognition and Emotion, 29(4), 736-746.
https://doi.org/10.1080/02699931.2014.932755
Paulhus, D. L., & Vazire, S. (2007). The self-report method. In R. W. Robins, R. C. Fraley, & R. F. Krueger (Eds.), Handbook of Research Methods in Personality Psychology (pp.
224-239). Guilford. Pavani, J.-B., Le Vigouroux, S., Kop, J.-L., Congard, A., & Dauvier, B. (2017). A network
approach to affect regulation dynamics and personality trait-induced variations:
Extraversion and neuroticism moderate reciprocal influences between affect and affect regulation strategies. European Journal of Personality, 31(4), 329–346.
https://doi.org/10.1002/per.2109 Pe, M. L., Raes, F., Koval, P., Brans, K., Verduyn, P., & Kuppens, P. (2013). Interference
resolution moderates the impact of rumination and reappraisal on affective
experiences in daily life. Cognition and Emotion, 27(3), 492–501. https://doi.org/10.1080/02699931.2012.719489
Peters, B. J., Overall, N. C., Cameron, L. D., Hammond, M. D., Low, R. S. T., & Girme, Y.
U. (2020). Do habitual emotional suppression measures predict response-focused situational suppression during social interactions? Emotion, 20, 1005-1019.
https://doi.org/10.1037/emo0000620 Peterson, R. A., & Brown, S. P. (2005). On the use of beta coefficients in meta-analysis.
Journal of Applied Psychology, 90(1), 175–181. https://doi.org/10.1037/0021-
9010.90.1.175
EMOTION REGULATION IN EVERYDAY LIFE 47
Quoidbach, J., Berry, E. V., Hansenne, M., & Mikolajczak, M. (2010). Positive emotion
regulation and well-being: Comparing the impact of eight savoring and dampening strategies. Personality and Individual Differences, 49(5), 368-373.
https://doi.org/10.1016/j.paid.2010.03.048 Rauthmann, J. F., Horstmann, K. T., & Sherman, R. A. (2018). Do self-reported traits and
aggregated states capture the same thing? A nomological perspective on trait-state
homomorphy. Social Psychological and Personality Science, 10(5), 596–611. https://doi.org/10.1177/1948550618774772
Robinson, M. D., & Clore, G. L. (2002). Belief and feeling: Evidence for an accessibility
model of emotional self-report. Psychological Bulletin, 128(6), 934–960. https://doi.org/10.1037/0033-2909.128.6.934
Rottenberg, J., & Hindash, A. C. (2015). Emerging evidence for emotion context insensitivity in depression. Current Opinion in Psychology, 4, 1-5.
https://doi.org/10.1016/j.copsyc.2014.12.025
Rozin, P., & Royzman, E. B. (2001). Negativity bias, negativity dominance, and contagion. Personality and Social Psychology Review, 5(4), 296–320.
https://doi.org/10.1207/S15327957PSPR0504_2 Schwartz, J. E., Neale, J., Marco, C., Shiffman, S. S., & Stone, A. A. (1999). Does trait
coping exist? A momentary assessment approach to the evaluation of traits. Journal of
Personality and Social Psychology, 77(2), 360–369. https://doi.org/10.1037//0022- 3514.77.2.360
Schwarz, N. (2012). Why researchers should think “real-time”: A cognitive rationale. In M.
R. Mehl & T. S. Conner (Eds.), Handbook of Research Methods for Studying Daily Life (pp. 22–42). Guilford.
Sheppes, G. (2020). Transcending the “good & bad” and “here & now” in emotion regulation: Costs and benefits of strategies across regulatory stages. Advances in
Experimental Social Psychology, 61, 185–236.
https://doi.org/10.1016/bs.aesp.2019.09.003
EMOTION REGULATION IN EVERYDAY LIFE 48
Sheppes, G., Scheibe, S., Suri, G., Radu, P., Blechert, J., & Gross, J. J. (2014). Emotion
regulation choice: A conceptual framework and supporting evidence. Journal of Experimental Psychology: General, 143(1), 163–181.
https://doi.org/10.1037/a0030831 Sheppes, G., Suri, G., & Gross, J. J. (2015). Emotion regulation and psychopathology.
Annual Review of Clinical Psychology, 11, 379–405. https://doi.org/10.1146/annurev-
clinpsy-032814-112739 Sun, J., & Vazire, S. (2019). Do people know what they’re like in the moment?
Psychological Science, 30(3), 405–414. https://doi.org/10.1177/0956797618818476
Suri, G., Whittaker, K., & Gross, J.J. (2015). Launching reappraisal: It’s less common than you might think. Emotion, 15, 73–77. https://doi.apa.org/doi/10.1037/emo0000011
Tamir, M. (2021). Effortful emotion regulation as a unique form of cybernetic control. Perspectives on Psychological Science, 16, 94–117.
https://doi.org/10.1177%2F1745691620922199
Tamir, M., Halperin, E., Porat, R., Bigman, Y. E., & Hasson, Y. (2019). When there’s a will, there’s a way: Disentangling the effects of goals and means in emotion regulation.
Journal of Personality and Social Psychology, 116(5), 795–816. https://doi.org/10.1037/pspp0000232
Trapnell, P. D., & Campbell, J. D. (1999). Private self-consciousness and the five-factor
model of personality: Distinguishing rumination from reflection. Journal of Personality and Social Psychology, 76(2), 284-304. https://doi.org/10.1037/0022-
3514.76.2.284
Treynor, W., Gonzalez, R., & Nolen-Hoeksema, S. (2003). Rumination reconsidered: A psychometric analysis. Cognitive Therapy and Research, 27(3), 247-259.
https://doi.org/10.1023/A:1023910315561 Troy, A. S., Saquib, S., Thal, J., & Ciuk, D. J. (2019). The regulation of negative and positive
affect in response to daily stressors. Emotion, 19(5), 751–763.
https://doi.org/10.1037/emo0000486
EMOTION REGULATION IN EVERYDAY LIFE 49
Verduyn, P., Delvaux, E., Van Coillie, H., Tuerlinckx, F., & Van Mechelen, I. (2009).
Predicting the duration of emotional experience: Two experience sampling studies. Emotion, 9, 83–91. https://doi.org/10.1037/a0014610
Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. Journal of Statistical Software, 36(3), 1-48. https://doi.org/10.18637/jss.v036.i03
Webb, T. L., Miles, E., & Sheeran, P. (2012a). Dealing with feeling: A meta-analysis of the
effectiveness of strategies derived from the process model of emotion regulation. Psychological Bulletin, 138(4), 775–808. https://doi.org/10.1037/a0027600
Webb, T. L., Schweiger Gallo, I., Miles, E., Gollwitzer, P. M., & Sheeran, P. (2012b).
Effective regulation of affect: An action control perspective on emotion regulation. European Review of Social Psychology, 23(1), 143–186.
https://doi.org/10.1080/10463283.2012.718134 Wrzus, C., & Roberts, B. W. (2017). Processes of personality development in adulthood: The
TESSERA framework. Personality and Social Psychology Review, 21(3), 253–277.
https://doi.org/10.1177/1088868316652279 Young, G., & Suri, G. (2020). Emotion regulation choice: A broad examination of external
factors. Cognition and Emotion, 34, 242–261. https://doi.org/10.1080/02699931.2019.1611544
Zyphur, M. J., & Oswald, F. L. (2015). Bayesian estimation and inference: A user’s guide.
Journal of Management, 41(2), 390–420. https://doi.org/10.1177/0149206313501200
Supplemental Materials
Table of Contents Supplementary Methods ...... 2 Supplementary Results ...... 5 Descriptive Statistics ...... 5 Supplemental Analyses ...... 7 Within-Person Variability as Moderator of Selection Correspondence...... 7 Average Within-Person Identification Slopes...... 8 Average Within-Person Implementation Slopes...... 9 Unique Effects of Selection, NA Identification, and NA Implementation on Global Self-Reports...... 9 Unique Effects of Selection, PA Identification, and PA Implementation on Global Self-Reports...... 11 Statistical Power For Meta-Analyses...... 12 References ...... 14
1
Supplementary Methods Tables S1 to S3 provide additional methodological details about the nine ESM and DD studies analyzed in the current report.
Table S1 Methodological Details of all Studies Daily Life Methods Final Age Data Sampling Compliance Previous Publication Study Name Females Context Sample M (SD) Exclusions protocol M (SD) N = 114 Negative Low compliance (n = 6) 1 DD x 7 days DD Study 1 35.23 (11.87) 50% 98% (7%) Kalokerinos et al. (2017) USA (MT) events Failed attn. check (n = 1) (T ~ 7) N = 153 1 DD x 21 days DD Study 2 18.70 (1.07) 56% Daily life none 89% (18%) Nezlek & Kuppens (2008) USA (UG) (T ~ 21) N = 176 9-10 ESMs x 21 days ESM Study 1 27.15 (9.03) 66% Daily life Low compliance (n = 3) 85% (12%) Grommisch et al. (2019) AUS (CO) (T ~ 200) N = 200 Transition 10 ESMs x 7 days ESM Study 2 18.32 (0.96) 55% Low compliance (n = 2) 87% (9%) Koval et al. (2015) BEL (UG) to university (T ~ 70) N = 46 10 ESMs x 7 days ESM Study 3 21.57 (3.88) 54% Daily life Low compliance (n = 4) 78% (13%) Brans et al. (2013), Study 1 AUS (UG) (T ~ 70) Low compliance (n = 1) N = 95 10 ESMs x 7 days ESM Study 4 19.06 (1.28) 62% Daily life Dropped out (n = 1) 92% (6%) Brans et al. (2013), Study 2 BEL (UG) (T ~ 70) Tech. error (n = 3) N = 101 Exam 10 ESMs x 9 days ESM-Event Study 1 18.64 (1.45) 86% none 91% (7%) Kalokerinos et al., (2019), Study 2 BEL (UG) results (T ~ 90) N = 100 Emotional 7 ESMs x 14 days ESM-Event Study 2 24.12 (6.87) 77% Low compliance (n = 4) 89% (10%) Dejonckheere et al., (2019) BEL (CO) events (T ~ 100) Low compliance (n = 11) N = 112 Emotional 8 ESMs x 7 days ESM-Event Study 3 21.20 (3.58) 68% Dropped out (n = 4) 84% (10%) Medland et al., (2020) AUS (UG) Events (T ~ 56) Tech. error (n = 5) Note. ESM = Experience Sampling Method; DD = Daily Diary; USA = United States of America; AUS = Australia; BEL = Belgium; UG = Undergraduate; CO = Community; MT = Mechanical Turk.
2
Table S2 Daily Self-Report Items Assessing Emotion Regulation Strategies in All Studies
Study Reappraisal item(s) Suppression item(s) Rumination item(s) Response Scale DD Study 1 I changed my perspective I suppressed the outward I ruminated or dwelled on 7-point scale: or the way I was thinking expression of my the event or my emotions. 1 = I did not do this at all about the event. emotions. 2 = I did this a little bit 7 = I did this very much DD Study 2 1. When I wanted to feel 1. When I was feeling 1. How much time did you 7-point scale: more positive emotion positive emotions, I was spend “ruminating” or 1= not at all (such as happiness or careful not to express dwelling on things that characteristic of me amusement), I changed them. happened to you for a long 7= very characteristic of what I was thinking about. 2. When I felt negative time afterward? me 2. When I wanted to feel emotions (such as sadness, 2. Today I played back less negative emotion, I nervousness, or anger), I over my mind how I acted changed what I was was careful not to express in a past situation. thinking about. them. 3. How much time did you 3. When I wanted to 3. I controlled my spend rethinking things control my emotions, I emotions by not that are over and done was not likely to change expressing them. with? the way I thought about the situation (reversed). ESM Study 1 1. I changed the way I was I was careful not to I thought over and over Slider scale: thinking about the express my emotions to again about my emotions. 0 = not at all situation. others. 100 = very much 2. I took a step back and looked at things from a different perspective. ESM Study 2 Have you looked at the Have you suppressed the 1. Have you ruminated Slider scale: cause of your feelings expression of your about something in the 0 = not at all from another perspective? feelings? past? 100 = very much 2. Have you ruminated about something in the future? ESM Study 3 I have changed the way I I have avoided expressing 1. I couldn’t stop thinking 6-point scale: think about what causes my emotions about my feelings. 0 = not at all my feelings 2. I have been focusing on 5 = very much my feelings. 3. I have been focusing on my problems. ESM Study 4 Have you looked at the Have you suppressed the Have you ruminated? Slider scale: cause of your feelings expression of your 1 = not at all from another perspective? feelings? 100 = very much ESM-Event Study 1 Have you looked at your Have you suppressed the Did you ruminate about 7-point scale: grades and the emotions outward expression of your grades? 0 = not at all that go with them from your emotions about your 1 = a little bit another perspective? grades? 6 = very much ESM-Event Study 2 To what extent did you try To what extent did you How much did you Slider scale: to change the way you suppress the outward ruminate on the event or 1 = not at all were thinking about the expression of your your emotions? 100 = very much event in order to change emotions? its emotional impact? ESM-Event Study 3 1. I thought of other ways 1. I made an effort to hide 1. I thought about the Slider scale: to interpret the situation. my feelings. emotional event again and 0 = not at all 2. I looked at the situation 2. I pretended I wasn't again. 100 = very much from several different upset. 2. I continually thought angles. about what was bothering me. Note. For all Daily Diary items, participants were instructed to rate their use of emotion regulation strategies in relation to “today”. In all ESM and ESM-Event studies, participants were instructed to rate their use of emotion regulation strategies “since the last survey”.
3
Table S3 Daily Self-Report Items Assessing Positive and Negative Affect in ESM and ESM-Event Studies Study Negative Affect item(s) Positive Affect item(s) Response Scale ESM Study 1 Angry, Sad, Stressed Happy, Relaxed, Confident Slider scale: 0 = not at all 100 = very much
ESM Study 2 Angry, Sad, Stressed, Anxious, Happy, Relaxed, Excited Slider scale: Depressed 0 = not at all 100 = very much ESM Study 3 Angry, Stressed, Anxious, Depressed Happy, Relaxed 6-point scale: 0 = not at all [e.g., angry] 5 = very [e.g., angry] ESM Study 4 Angry, Sad, Anxious, Depressed Happy, Relaxed Slider scale: 0 = not at all 100 = very much ESM-Event Study 1 Angry, Sad, Stressed, Anxious, Happy, Proud, Content, Relief Slider scale: Ashamed, Disappointed 0 = not at all 100 = very much ESM-Event Study 2 Angry, Sad, Stressed Happy, Relaxed Slider scale: 0 = not at all 100 = very much ESM-Event Study 3 What is the strongest negative emotion How are you feeling right now? Slider scale: you have experienced in the last hour? Negative Affect: (select one: angry/frustrated, 0 = Not at all, I barely sad/disappointed, anxious/stressed, noticed embarrassed/self-conscious) 100 = Very intense
How intense was the negative emotion? Positive Affect: -10 = very negative 0 = neutral +10 = very positive
Note. All affect items assessed momentary feelings (e.g., “How angry do you feel at the moment?”) except the Negative Affect item in ESM-Event Study 3, which assessed emotional intensity “in the last hour”.
4 Supplementary Results Descriptive Statistics Below, we report descriptive statistics and reliability estimates for global self-reports of emotion regulation strategies (Table S4) and daily self-reports of emotion regulation strategies (Table S5) across all nine studies. We also report the descriptive statistics and reliabilities for daily self-reports of negative (NA) and positive affect (PA) in the seven ESM and ESM-Event studies, in which affect ratings were analyzed (Table S6).
Table S4 Descriptive Statistics and Reliabilities for Global Self-Report Measures of Emotion Regulation
Reappraisal Suppression Rumination Study M SD a M SD a M SD a DD Study 1 4.99 1.18 .93 3.89 1.38 .81 2.02 0.79 .87 DD Study 2 4.46 1.08 .86 3.44 1.27 .79 4.27 1.25 .94 ESM Study 1 4.99 1.08 .86 3.86 1.42 .82 2.28 0.66 .73 ESM Study 2 4.63 0.93 .75 3.43 1.24 .80 2.07 0.61 .74 ESM Study 3 4.85 0.99 .83 3.94 1.21 .78 3.56 0.70 .91 ESM Study 4 4.42 0.88 .71 2.97 1.28 .82 2.11 0.62 .64 ESM-Event Study 1 4.76 0.86 .76 3.20 1.17 .81 2.28 0.62 .76 ESM-Event Study 2 4.85 0.90 .78 3.08 1.22 .79 2.19 0.63 .75 ESM-Event Study 3 4.95 1.00 .81 3.68 1.44 .85 2.37 0.55 .90 Note. Global reappraisal and suppression were assessed using Gross & John’s (2003) ERQ, rated on 1 to 7 scale. Global rumination was assessed with the brooding subscale of Treynor et al.’s (2003) version of the RRS rated on 1 to 4 scale in all studies except Daily Diary Study 1 and ESM Study 3, in which rumination was assessed with the rumination subscale of Trapnell and Campbell’s (1999) rumination-reflection questionnaire (with ratings from 1 to 5).
5
Table S5 Descriptive Statistics for Daily Self-Report Measures of Emotion Regulation
Reappraisal Suppression Rumination
M SDW SDB ICC items !W !B M SDW SDB ICC items !W !B M SDW SDB ICC items !W !B DD Study 1 2.70 1.62 1.07 .30 1 — — 3.35 1.84 1.11 .27 1 — — 3.14 1.65 1.18 .34 1 — — DD Study 2 3.60 0.95 1.28 .64 3 .79 .97 2.99 0.93 1.20 .62 3 .70 .93 3.53 1.18 1.21 .51 1 .85 .98 ESM Study 1 40.30 19.31 21.81 .56 2 .59 .99 38.55 24.64 22.76 .46 1 — — 36.70 23.26 21.82 .47 1 — — ESM Study 2 15.23 14.40 10.74 .36 1 — — 20.12 17.62 17.32 .49 1 — — 24.81 16.33 14.02 .42 2 .40 .81 ESM Study 3 2.11 0.93 0.89 .48 1 — — 2.47 1.17 1.00 .42 1 — — 2.35 0.89 0.84 .47 3 .74 .98 ESM Study 4 18.27 15.29 11.68 .37 1 — — 23.94 18.92 15.65 .41 1 — — 27.29 20.10 17.04 .42 1 — — ESM-Event Study 1 0.54 0.93 0.79 .42 1 — — 0.61 0.92 0.93 .50 1 — — 0.84 1.24 0.93 .36 1 — — ESM-Event Study 2 33.65 28.63 16.68 .25 1 — — 30.97 27.37 18.04 .30 1 — — 28.07 28.01 13.97 .20 1 — — ESM-Event Study 3 41.29 19.26 20.66 .54 2 .69 .99 36.78 21.58 18.57 .43 2 .71 .99 39.65 20.84 18.83 .45 2 .75 .99 Note. Daily self-reports of emotion regulation strategies were assessed using scales from 0 to 5 (ESM Study 3), 1 to 7 (DD Studies 1 & 2; ESM-Event Study 1), 0 to 100 (ESM Studies 1 & 2; ESM- Event Study 3), or 1 to 100 (ESM Study 4; ESM-Event Study 2). Multilevel Cronbach’s alphas were estimated following Geldhof et al. (2014).
Table S6 Descriptive Statistics for Daily Self-Report Measures of Negative and Positive Affect Negative Affect Positive Affect
M SDW SDB ICC Items !W !B M SDW SDB ICC Items !W !B ESM Study 1 22.15 14.24 13.54 .47 3 .62 .89 63.11 15.54 13.35 .42 3 0.72 .93 ESM Study 2 14.32 10.41 8.38 .39 5 .72 .93 56.87 16.83 9.82 .25 3 0.73 .90 ESM Study 3 2.10 0.76 0.78 .51 4 .74 .95 4.07 0.97 0.65 .30 2 0.62 .91 ESM Study 4 15.65 10.99 10.80 .49 4 .74 .92 57.27 17.91 13.18 .35 2 .71 .90 ESM-Event Study 1 27.58 14.14 18.20 .62 6 .85 .96 43.24 17.73 23.25 .63 4 .91 .98 ESM-Event Study 2 15.99 13.95 9.21 .30 3 .62 .83 62.36 18.67 11.04 .26 2 .72 .91 ESM-Event Study 3 39.38 23.30 14.28 .27 1 — — 2.91 3.83 2.89 .36 1 — — Note. Multilevel Cronbach’s alphas were estimated following Geldhof et al. (2014).
6 Supplemental Analyses Within-Person Variability as Moderator of Selection Correspondence. Table S7 contains results of models testing whether selection correspondence was moderated by individual differences in within-person variability in daily emotion regulation. In each model, global self-reports of an emotion regulation strategy were simultaneously regressed onto the mean, within-person variance (estimated as a latent residual variance), and latent mean*variability interaction of daily self-reports for the same regulation strategy. Meta-analyses showed no reliable evidence across studies that within-person variability in daily emotion regulation moderated selection correspondence.
Table S7 Standardized Regression Coefficients for Mean, Within-Person Variability and the Mean*Variability Interaction of Daily Self-Reported Emotion Regulation Predicting Global Self-Reported Emotion Regulation Daily Self-Reports Mean Within-Person Var Mean * Var Global Self-Reports Est. 95% CI Est. 95% CI Est. 95% CI Reappraisal DD Study 1 (N = 114) .29 .04 to .54 -.28 -.67 to .12 .17 -.10 to .42 DD Study 2 (N = 153) .18 .01 to .35 .66 .41 to .83 -.17 -.25 to -.08 ESM Study 1 (N = 176) .28 -.37 to .74 .02 -.20 to .21 -.02 -.11 to .10 ESM Study 2 (N = 200) -.03 -.65 to .56 .17 -.03 to .35 .00 -.15 to .15 ESM Study 3 (N = 46) .06 -.32 to .45 -.13 -.71 to .48 -.02 -.39 to .34 ESM Study 4 (N = 95) -.12 -.77 to .69 -.08 -.34 to .21 .06 -.14 to .21 ESM-Event Study 1 (N = 101) -.19 -.42 to .06 .13 -.15 to .38 -.31 -.70 to .05 ESM-Event Study 2 (N = 100) .64 -.74 to .93 .12 -.18 to .29 -.06 -.09 to .07 ESM-Event Study 3 (N = 112) .04 -.62 to .68 -.18 -.46 to .17 .07 -.08 to .18 Meta-Analysisa (N = 1097) .14 -.04 to .33 .07 -.14 to .28 -.03 -.13 to 06 Suppression DD Study 1 (N = 114) .18 -.18 to .46 .18 -.24 to .55 -.04 -.22 to .17 DD Study 2 (N = 153) .56 .39 to .71 .00 -.29 to .27 -.01 -.16 to .15 ESM Study 1 (N = 176) .66 .25 to .86 .06 -.13 to .20 -.06 -.12 to .04 ESM Study 2 (N = 200) -.12 -.69 to .62 -.18 -.35 to .00 .08 -.08 to .19 ESM Study 3 (N = 46) .28 -.04 to .60 -.34 -.83 to .27 .16 -.25 to .51 ESM Study 4 (N = 95) .50 -.53 to .92 .05 -.17 to .26 -.09 -.25 to .22 ESM-Event Study 1 (N = 101) -.21 -.51 to .09 .20 -.07 to .45 .53 .03 to 1.03 ESM-Event Study 2 (N = 100) .51 -.51 to .94 -.07 -.30 to .14 -.04 -.14 to .15 ESM-Event Study 3 (N = 112) .71 .20 to .91 .10 -.16 to .27 -.10 -.16 to .01 Meta-Analysisa (N = 1097) .40 .14 to .65 .01 -.09 to .11 .05 -.09 to .19 Rumination Daily Diary Study 1 (N = 114) .14 -.31 to .51 .17 -.32 to .58 .05 -.14 to .27 Daily Diary Study 2 (N = 153) .55 .40 to .69 .04 -.40 to .45 .00 -.14 to .16 ESM Study 1 (N = 176) .73 .47 to .88 .14 .01 to .25 -.08 -.12 to .01 ESM Study 2 (N = 200) .22 -.54 to .77 .10 -.12 to .28 .03 -.07 to .14 ESM Study 3 (N = 46) .31 -.11 to .66 -.60 -1.07 to .16 .27 -.07 to .53 ESM Study 4 (N = 95) -.24 -.90 to .68 .05 -.17 to .29 .12 -.11 to .26 ESM-Event Study 1 (N = 101) .26 .02 to .49 -.04 -.31 to .21 .06 -.20 to .33 ESM-Event Study 2 (N = 100) -.23 -.90 to .74 .08 -.15 to .35 .06 -.07 to .13 ESM-Event Study 3 (N = 112) .77 .31 to .94 .23 .00 to .39 -.10 -.15 to .00 Meta-Analysisa (N = 1097) .34 .04 to .64 .03 -.12 to .18 .02 -.04 to .08 Note. Estimates in bold have 95% Bayesian credible intervals (CIs) that do not include zero. a Meta-analytic correlations are Fisher z-transformed and values within the ‘95% CI’ column are frequentist 95% confidence intervals
7 Average Within-Person Identification Slopes. Tables S8 and S9 contain estimates of
average within-person NA and PA identification slopes. These estimates are taken from the same
models as those reported in the manuscript (see Figure 1 for model diagram).
Table S8 Average Within-Person Negative Affect (NA) Identification Slopes
NA Identification Slope (NAt-1 → ERt)
#$"% &'%% #'( !!" !!" !!" Study Est. 95% CI Est. 95% CI Est. 95% CI ESM Study 1 (N = 176) .03 .02 to .04 .05 .03 to .06 .07 .06 to .09 ESM Study 2 (N = 200) .07 .05 to .09 .07 .06 to .09 .12 .10 to .13 ESM Study 3 (N = 46) .06 .01 to .10 .06 .01 to .11 .09 .04 to .15 ESM Study 4 (N = 95) .08 .05 to .10 .07 .04 to .09 .15 .12 to .18 ESM-Event Study 1 (N = 101) .10 .08 to .12 .10 .09 to .13 .17 .15 to .19 ESM-Event Study 2 (N = 100) .04 .02 to .06 .01 -.01 to .04 .08 .06 to .10 ESM-Event Study 3 (N = 112) .00 -.03 to .03 .00 -.04 to .02 .07 .04 to .10 Meta-Analysisa (N = 830) .05 -.02 to .12 .05 -.02 to .12 .11 .04 to .18 Note. CI = Bayesian credibility intervals; ER = emotion regulation strategy; NA = negative affect; a Meta-analytic correlations are Fisher z-transformed and values within the ‘95% CI’ column are frequentist confidence intervals
Table S9 Average Within-Person Positive Affect (PA) Identification Slopes
PA Identification Slope (PAt-1 → ERt)
#$"% &'%% #'( !%" !%" !%" Study Est. 95% CI Est. 95% CI Est. 95% CI ESM Study 1 (N = 176) .00 -.01 to .01 -.04 -.05 to -.02 -.04 -.05 to -.03 ESM Study 2 (N = 200) -.01 -.03 to .00 -.07 -.08 to -.05 -.10 -.12 to -.08 ESM Study 3 (N = 46) .00 -.05 to .04 -.01 -.06 to .03 -.03 -.08 to .01 ESM Study 4 (N = 95) -.04 -.07 to -.01 -.06 -.08 to -.03 -.12 -.15 to .10 ESM-Event Study 1 (N = 101) -.05 -.07 to -.03 -.06 -.08 to -.02 -.10 -.12 to -.07 ESM-Event Study 2 (N = 100) -.04 -.06 to -.02 -.03 -.06 to -.01 -.09 -.11 to -.06 ESM-Event Study 3 (N = 112) .02 -.01 to .05 .02 -.01 to .05 -.07 -.10 to -.04 Meta-Analysisa (N = 830) -.02 -.08 to .05 -.04 -.11 to .03 -.08 -.15 to -.01 Note. CI = Bayesian credibility intervals; ER = emotion regulation strategy; PA = positive affect; a Meta-analytic correlations are Fisher z-transformed and values within the ‘95% CI’ column are frequentist confidence intervals
8 Average Within-Person Implementation Slopes. Tables S10 and S11 contain
estimates of average within-person NA and PA implementation slopes. These estimates are
taken from the same models as those reported in the manuscript (see Figure 1 for model
diagram).
Table S10 Average Within-Person Negative Affect (NA) Implementation Slopes
NA Implementation Slope (ERt → NAt)
!" !" !" "#$"% "&'%% "#'( Study Est. 95% CI Est. 95% CI Est. 95% CI ESM Study 1 (N = 176) -.07 -.08 to -.06 .08 .07 to .10 .18 .16 to .19 ESM Study 2 (N = 200) .06 .04 to .09 .19 .14 to .20 .21 .20 to .24 ESM Study 3 (N = 46) -.01 -.07 to .04 .09 .04 to .14 .36 .30 to .42 ESM Study 4 (N = 95) .04 .01 to .07 .17 .14 to .21 .33 .30 to .36 ESM-Event Study 1 (N = 101) -.09 -.12 to -.06 .02 -.01 to .05 .13 .11 to .16 ESM-Event Study 2 (N = 100) .06 .04 to .09 .09 .06 to .12 .29 .26 to .32 ESM-Event Study 3 (N = 112) .00 -.03 to .04 .08 .05 to .11 .32 .29 to .35 Meta-Analysisa (N = 830) .00 -.07 to .07 .11 .04 to .18 .25 .18 to .32 Note. CI = Bayesian credibility intervals; ER = emotion regulation strategy; NA = negative affect; a Meta-analytic correlations are Fisher z-transformed and values within the ‘95% CI’ column are frequentist confidence intervals
Table S11 Average Within-Person Positive Affect (PA) Implementation Slopes
PA Implementation Slope (ERt → PAt)
%" %" %" "#$"% "&'%% "#'( Study Est. 95% CI Est. 95% CI Est. 95% CI ESM Study 1 (N = 176) .14 .13 to .16 -.06 -.08 to -.05 -.12 -.14 to -.11 ESM Study 2 (N = 200) .06 .04 to .09 -.08 -.10 to -.06 -.18 -.21 to -.16 ESM Study 3 (N = 46) .08 .02 to .13 -.06 -.11 to -.01 -.19 -.25 to -.14 ESM Study 4 (N = 95) .06 .03 to .10 -.10 -.13 to -.07 -.24 -.27 to -.21 ESM-Event Study 1 (N = 101) .11 .08 to .14 .01 -.02 to .04 -.05 -.07 to -.03 ESM-Event Study 2 (N = 100) -.02 -.05 to .00 -.08 -.11 to -.05 -.23 -.26 to -.20 ESM-Event Study 3 (N = 112) .05 .02 to .09 -.01 -.04 to .02 -.20 -.24 to -.17 Meta-Analysisa (N = 830) .07 .005 to .14 -.06 -.13 to .01 -.17 -.24 to -.11 Note. CI = Bayesian credibility intervals; ER = emotion regulation strategy; PA = positive affect; a Meta-analytic correlations are Fisher z-transformed and values within the ‘95% CI’ column are frequentist confidence intervals
Unique Effects of Selection, NA Identification, and NA Implementation on
Global Self-Reports. Table S12 contains estimates of selection, identification, and
implementation correspondence from MSEM models, in which scores on each global
emotion regulation measure were simultaneously regressed onto the selection, NA
9
identification and NA implementation parameters modeled using daily self-reports of the
corresponding regulation strategy.
Table S12 Standardized Regression Weights for Daily Strategy Selection, NA Identification, and NA Implementation Predicting Global Self-Reports of each Emotion-Regulation Strategy
Simultaneous Predictors Selection Identification Implementation
!"#$ !"#$ %# ! "%# #!"#$ Outcome variable est. 95% CI est. 95% CI est. 95% CI Global Reappraisal ESM Study 1 (N = 176) .21 .03 to .40 .09 -.17 to .34 -.05 -.30 to .20 ESM Study 2 (N = 200) .01 -.16 to .18 .12 -.10 to .34 -.16 -.40 to .07 ESM Study 3 (N = 46) .05 -.43 to .54 -.35 -.93 to .43 .12 -.46 to .68 ESM Study 4 (N = 95) -.07 -.38 to .22 .42 .07 to .74 -.10 -.49 to .28 ESM-Event Study 1 (N = 101) -.20 -.42 to .02 .20 -.04 to .44 .03 -.24 to .28 ESM-Event Study 2 (N = 100) -.08 -.34 to .17 -.19 -.52 to .16 -.14 -.47 to .23 ESM-Event Study 3 (N = 112) .45 .14 to .79 .27 -.12 to .63 .22 -.25 to .66 Meta-Analysisa (N = 830) .06 -.11 to .23 .09 -.10 to .29 -.02 -.13 to .08 &'$$ &'$$ %# ! "%# #&'$$ est. 95% CI est. 95% CI est. 95% CI Global Suppression ESM Study 1 (N = 176) .47 .30 to .63 -.05 -.35 to .23 .04 -.26 to .33 ESM Study 2 (N = 200) .21 .06 to .37 .00 -.21 to .20 .12 -.08 to .32 ESM Study 3 (N = 46) .16 -.28 to .57 -.59 -.98 to -.07 -.02 -.50 to .44 ESM Study 4 (N = 95) .20 -.04 to .45 -.23 -.55 to .12 -.13 -.42 to .16 ESM-Event Study 1 (N = 101) .12 -.11 to .33 .13 -.13 to .37 .07 -.26 to .44 ESM-Event Study 2 (N = 100) .41 .11 to .70 -.47 -.82 to -.08 .19 -.22 to .60 ESM-Event Study 3 (N = 112) .23 -.01 to .45 .18 -.27 to .61 -.16 -.48 to .16 Meta-Analysisa (N = 830) .28 .16 to .39 -.15 -.39 to .08 .02 -.07 to .12 !'( !'( %# ! "%# #!'( est. 95% CI est. 95% CI est. 95% CI Global Rumination ESM Study 1 (N = 176) .50 .36 to .63 -.07 -.28 to .14 .23 .03 to .44 ESM Study 2 (N = 200) .46 .34 to .58 -.02 -.20 to .16 .05 -.11 to .21 ESM Study 3 (N = 46) .18 -.27 to .60 -.12 -.78 to .67 .45 -.06 to .87 ESM Study 4 (N = 95) .31 .07 to .54 -.01 -.35 to .32 .12 -.17 to .40 ESM-Event Study 1 (N = 101) .29 .07 to .48 -.19 -.42 to .05 .06 -.18 to .31 ESM-Event Study 2 (N = 100) .42 .18 to .65 -.30 -.59 to .03 .08 -.22 to .36 ESM-Event Study 3 (N = 112) .28 .07 to .49 -.16 -.49 to .21 .00 -.29 to .31 Meta-Analysisa (N = 830) .40 .30 to .49 -.11 -.20 to -.03 .12 .03 to .21 Note. Shaded cells contain correlations representing identification correspondence for each emotion regulation strategy; Estimates in bold have 95% Bayesian credible intervals (CIs) that do not include zero. a Meta-analytic correlations are Fisher z-transformed and values within the ‘95% CI’ column are frequentist 95% confidence intervals
10
Unique Effects of Selection, PA Identification, and PA Implementation on
Global Self-Reports. Table S13 contains estimates of selection, identification, and
implementation correspondence from MSEM models, in which scores on each global
emotion regulation measure were simultaneously regressed onto the selection, PA
identification and PA implementation parameters modeled using daily self-reports of the
corresponding regulation strategy.
Table S13 Standardized Regression Weights for Daily Strategy Selection, PA Identification, and PA Implementation Predicting Global Self-Reports of each Emotion-Regulation Strategy
Daily Emotion Regulation Process
Selection Identification Implementation
#$"% #$"% %" # !%" "#$"% Global Self-Reports est. 95% CI est. 95% CI est. 95% CI Reappraisal ESM Study 1 (N = 176) .26 .05 to .47 -.31 -.55 to -.06 .12 -.17 to .38 ESM Study 2 (N = 200) .06 -.09 to .22 -.03 -.29 to .24 -.16 -.41 to .11 ESM Study 3 (N = 46) -.01 -.47 to .45 .01 -.61 to .64 -.20 -.75 to .46 ESM Study 4 (N = 95) -.07 -.48 to .26 -.45 -.93 to .01 .44 -.15 to .98 ESM-Event Study 1 (N = 101) -.28 -.48 to -.06 -.06 -.30 to .17 -.24 -.47 to .02 ESM-Event Study 2 (N = 100) -.08 -.40 to .24 .09 -.30 to .47 .09 -.39 to .6 ESM-Event Study 3 (N = 112) .35 .15 to .56 .10 -.21 to .4 -.05 -.34 to .23 Meta-Analysisa (N = 830) .04 -.13 to .21 -.10 -.27 to .06 .01 -.18 to .19 &'%% &'%% %" # !%" "&'%% est. 95% CI est. 95% CI est. 95% CI Suppression ESM Study 1 (N = 176) .53 .25 to .82 .16 -.14 to .44 -.17 -.60 to .26 ESM Study 2 (N = 200) .27 .07 to .47 .17 -.09 to .41 -.15 -.38 to .10 ESM Study 3 (N = 46) .10 -.35 to .56 .53 -.01 to .97 .25 -.34 to .78 ESM Study 4 (N = 95) .34 .07 to .60 .19 -.26 to .61 -.21 -.60 to .18 ESM-Event Study 1 (N = 101) .12 -.10 to .35 .00 -.24 to .26 .07 -.25 to .37 ESM-Event Study 2 (N = 100) .39 .13 to .66 .36 -.07 to .74 .06 -.26 to .38 ESM-Event Study 3 (N = 112) .22 -.01 to .45 .45 .07 to .76 -.15 -.48 to .18 Meta-Analysisa (N = 830) .32 .18 to .44 .26 .13 to .41 -.07 -.18 to .04 #'( #'( %" # !%" "#'( est. 95% CI est. 95% CI est. 95% CI Rumination ESM Study 1 (N = 176) .57 .41 to .71 .21 -.04 to .48 -.37 -.64 to -.12 ESM Study 2 (N = 200) .44 .29 to .58 -.09 -.32 to .15 -.01 -.22 to .19 ESM Study 3 (N = 46) .17 -.30 to .63 .34 -.32 to .86 .02 -.59 to .61 ESM Study 4 (N = 95) .30 .03 to .55 -.24 -.66 to .25 -.02 -.35 to .33 ESM-Event Study 1 (N = 101) .30 .09 to .51 .03 -.20 to .27 -.09 -.35 to .18 ESM-Event Study 2 (N = 100) .37 .14 to .58 .07 -.25 to .40 -.06 -.36 to .23 ESM-Event Study 3 (N = 112) .28 .07 to .48 .22 -.09 to .51 -.09 -.34 to .17 Meta-Analysisa (N = 830) .39 .27 to .51 .07 -.07 to .21 -.10 -.22 to .01 Note. Shaded cells contain correlations representing identification correspondence for each emotion regulation strategy; Estimates in bold have 95% Bayesian credible intervals (CIs) that do not include zero. a Meta-analytic correlations are Fisher z-transformed and values within the ‘95% CI’ column are frequentist 95% confidence intervals 11
Statistical Power For Meta-Analyses. We calculated post-hoc power for our main random effects meta-analyses following Valentine et al. (2010), using an adapted R-script made available by Quintana (2017). This revealed that our meta-analyses including all studies
(k = 9) had 80% power to detect effects as small as r = .07, assuming small between-study heterogeneity; r = .08, assuming moderate between-study heterogeneity; and r = .12, assuming large between-study heterogeneity (see Valentine et al., 2010). For meta-analyses including only the ESM and ESM-Event studies (k = 7), we had 80% power to detect effects as small as r = .08, assuming small between-study heterogeneity; r = .10 assuming moderate between-study heterogeneity; and r = .14, assuming large between-study heterogeneity.
Given that between-study heterogeneity in effect sizes was large in some cases, we calculated the observed power for the two largest non-significant meta-analytic effects obtained in our main analyses (i.e., those reported in the main text, rather than the supplemental materials). These were (i) the association between global suppression and the
NA identification slope for state suppression (rmeta = –.11, 95% CI [-.24 to .03]), for which observed power was 63%; and (ii) the association between global rumination and the PA implementation slope for state reappraisal (rmeta = .14, 95% CI [–.01 to .30]), for which our observed power was 77%. Thus, our meta-analyses were underpowered to detect some potentially significant effects due to high levels of between-study heterogeneity.
Finally, we calculated the observed power for all meta-analytic effects in our main analyses that differed significantly from zero, which we report in Table S14 (below). We report absolute meta-analytic effect sizes, r, and I2 estimates obtained using the metafor R- package (Viechtbauer, 2010). These meta-analytic effect sizes were converted to Cohen’s ds using the effectsize R-package (Ben-Shachar et al., 2020). Finally, we calculated the observed power for each meta-analysis by adapting Quintana’s (2017) R-script. As shown in the table below, observed power ranged between .90 and 1 for most (15 out of 18) meta-analytic effects. 12
Table S14 Post-Hoc (observed) Power for all Statistically Significant Meta-Analyses Reported in Main Manuscript Source Table Effect |r| I2 Cohen's d k Average N Power
!" Table 4 Global Reappraisal with "#$"% .07 0.00% 0.14 7 118.57 .83 &'%% Table 3 Global Suppression with !%" .11 55.6% 0.23 7 118.57 .87 Table 1 Global Reappraisal with ##$"% .14 79.5% 0.29 9 121.89 .87 &'%% Table 2 Global Rumination with !!" .08 0.00% 0.16 7 118.57 .90 #'( Table 2 Global Suppression with !!" .08 0.00% 0.16 7 118.57 .92 #'( Table 3 Global Reappraisal with !%" .09 0.00% 0.17 7 118.57 .94 !" Table 4 Global Rumination with "#$"% .14 64.7% 0.29 7 118.57 .94 #'( Table 2 Global Rumination with !!" .09 0.0% 0.18 7 118.57 .96 #$"% Table 3 Global Suppression with !%" .15 64.6% 0.30 7 118.57 .96 Table 1 Global Suppression with ##'( .13 58.1% 0.26 9 121.89 .97 Table 1 Global Rumination with ##$"% .13 63.2% 0.27 9 121.89 .97 !" Table 4 Global Rumination with "#'( .10 1.00% 0.20 7 118.57 .98 #$"% Table 2 Global Suppression with !!" .15 0.00% 0.30 7 118.57 ~1.00 #$"% Table 3 Global Rumination with !%" .17 27.2% 0.35 7 118.57 ~1.00 #$"% Table 2 Global Rumination with !!" .19 44.0% 0.39 7 118.57 ~1.00 Table 1 Global Rumination with #&'%% .25 64.1% 0.52 9 121.89 ~1.00 Table 1 Global Suppression with #&'%% .30 70.4% 0.64 9 121.89 ~1.00 Table 1 Global Rumination with ##'( .40 60.4% 0.86 9 121.89 ~1.00 Note. |r| = absolute value of meta-analytic correlation. I2 = proportion of total variation in effect sizes attributable to between- study variance; Cohen’s d = meta-analytic correlation converted to Cohen’s d; k = number of effects in meta-analysis; Average N = average sample size per study in meta-analysis.
13
References Ben-Shachar, Makowski & Lüdecke (2020). Compute and interpret indices of effect size.
CRAN. https://github.com/easystats/effectsize.
Brans, K., Koval, P., Verduyn, P., Lim, Y. L., & Kuppens, P. (2013). The regulation of
negative and positive affect in daily life. Emotion, 13(5), 926–939.
https://doi.org/10.1037/a0032400
Dejonckheere, E., Kalokerinos, E. K., Bastian, B., & Kuppens, P. (2019). Poor emotion
regulation ability mediates the link between depressive symptoms and affective
bipolarity. Cognition and Emotion, 33(5), 1076–1083.
https://doi.org/10.1080/02699931.2018.1524747
Geldhof, G. J., Preacher, K. J., & Zyphur, M. J. (2014). Reliability estimation in a multilevel
confirmatory factor analysis framework. Psychological Methods, 19(1), 72–91.
https://doi.org/10.1037/a0032138
Grommisch, G., Koval, P., Hinton, J. D. X., Gleeson, J., Hollenstein, T., Kuppens, P., &
Lischetzke, T. (2019). Modeling individual differences in emotion regulation
repertoire in daily life with multilevel latent profile analysis. Emotion, 20(8), 1462–
1474. https://doi.org/10.1037/emo0000669
Kalokerinos, E. K., Erbas, Y., Ceuelemans, E., & Kuppens, P. (2019). Differentiate to
regulate: Low negative emotion differentiation is associated with ineffective use but
not selection of emotion-regulation strategies. Psychological Science, 30(6), 863-879.
https://doi.org/10.1177/0956797619838763
Kalokerinos, E. K., Tamir, M., & Kuppens, P. (2017). Instrumental motives in negative
emotion regulation in daily life: Frequency, consistency, and predictors. Emotion,
17(4), 648-657. https://doi.org/10.1037/emo0000269
14
Koval, P., Brose, A., Pe, M. L., Houben, M., Erbas, Y., Champagne, D., & Kuppens, P.
(2015). Emotional inertia and external events: The roles of exposure, reactivity, and
recovery. Emotion, 15(5), 625–636. https://doi.org/10.1037/emo0000059
Medland, H., De France, K., Hollenstein, T., Mussoff, D., & Koval, P. (2020). Regulating
emotion Systems in everyday life: Reliability and validity of the RESS-EMA scale.
European Journal of Psychological Assessment, 36(3), 437–446.
https://doi.org/10.1027/1015-5759/a000595
Nezlek, J. B., & Kuppens, P. (2008). Regulating positive and negative emotions in daily life.
Journal of Personality, 76(3), 561–580. https://doi.org/10.1111/j.1467-
6494.2008.00496.x
Quintana, D. (2017, July 29). How to calculate statistical power for your meta-analysis.
Towards Data Science. https://towardsdatascience.com/how-to-calculate-statistical-
power-for-your-meta-analysis-e108ee586ae8
Valentine, J. C., Pigott, T. D., & Rothstein, H. R. (2010). How many studies do you need? A
primer on statistical power for meta-analysis. Journal of Educational and Behavioral Statistics, 35(2), 215-247. https://doi.org/10.3102/1076998609346961
Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. Journal of
Statistical Software, 36(3), 1-48. https://doi.org/10.18637/jss.v036.i03
15